U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-05-137
Date: July 2006

Evaluation of Safety, Design, and Operation of Shared-Use Paths

Final Report

PDF Version (1172 KB)

PDF files can be viewed with the Acrobat® Reader®





Shared–use paths are paved, off–street travel ways designed to serve nonmotorized travelers. Across the United States, bicyclists are typically the most common users of shared–use paths. However, in many places, shared–use paths are frequently used by pedestrians, inline skaters, roller skaters, skateboarders, wheelchair users, and users of many other modes. In many places, Segway® Human Transporters (Segway HT) are allowed on shared–use paths and blur the line between motorized and nonmotorized modes. In the United States, there are very few paths limited exclusively to bicyclists. Most off–street paths in this country fall into the shared–use path category. We should note that the term "trail" is used interchangeably with the term "shared–use path" in this report.

Most shared–use paths in the United States are constructed to provide recreational opportunities. Some are also intended to serve commuters. Shared–use paths are also very common on university campuses because motor vehicle traffic and parking are often heavily restricted.

Gaining Popularity

Shared–use paths are gaining popularity in two different ways in recent years in the United States. First, new path segments are opening across the United States all of the time. Whether they are in old railroad rights–of–way, on creekside and riverside flood plains, on the banks of reservoirs and lakes, or in rights–of–way set aside by developers, almost every medium–sized and large urban area in the United States has some shared–use paths and has plans for more. Funding for path construction is being provided by Federal, State, and local governments and by private sources. There is no sign that the pace of construction of new shared–use paths is slowing.

The greatest testimony to the success of the trails movement in the United States is the enormous amount of use they have attracted. Some urban trails attract thousands of users per hour during peak periods. Many trails are experiencing morning rush hours on weekdays and traffic jams on weekend afternoons. Trail managers in many parts of the country are becoming increasingly concerned about user conflicts and injuries. Some are also concerned that potential users are deciding not to use a trail because of crowding.


During the design of every shared–use path, someone eventually asks how wide should a pathway be. That question nearly always raises even more questions: What types of users can we reasonably expect? When will we need to widen the path? Do we need to separate different types of users from each other? These are very difficult questions for designers. They face that classic design dilemma of overbuilding versus obsolescence. If the designer specifies a trail wider than future use justifies, money is wasted that could have otherwise gone to construct more miles of trail elsewhere. If the designer specifies a trail that proves to be too narrow for the future volume and mix of users, there will be more user conflicts and collisions, greater unhappiness among users, and the need to consider expensive trail widening.

At this time, conventional design manuals do little to help designers resolve their dilemmas. The 1999 American Association of State Highway and Transportation Officials (AASHTO) Guide for the Development of Bicycle Facilities states, "Under most conditions, a recommended paved width for a two–directional shared–use path is 3 meters [m] (10 feet [ft])... Under certain conditions, it may be necessary or desirable to increase the width of a shared–use path to 3.6 m (12 ft) or even 4.2 m (14 ft), due to substantial use by bicycles, joggers, skaters, and pedestrians."(1) No further guidance is given to determine what specific levels of use–or mixture of uses–warrant a wider pathway or a separation of users.

Existing Level of Service (LOS) Method

Versions of the Highway Capacity Manual (HCM) prior to the year 2000 contained no help for trail designers.(2) There were no quality–of–service procedures for shared–use paths.

A recent research effort, conducted by several of the authors of this report and sponsored by the Federal Highway Administration (FHWA), attempted to fill this information gap. Rouphail, et al., recommended an analytical procedure to determine the LOS for bicyclists on shared off–street paths for inclusion in the 2000 edition of the HCM. (23) The Transportation Research Board (TRB) Highway Capacity and Quality of Flow Committee, which oversees the HCM, agreed with the recommendation and the 2000 edition contained the procedure. (4) Rouphail, et al., adapted the procedure that was originally developed by Hein Botma (also a member of this research team), based on simulations and field studies from The Netherlands. (2,5)

Botma's model, which is discussed in depth in the literature review in chapter 2 of this report, is based on fundamental traffic–flow theory. The Botma model works much like a model of vehicular traffic on a roadway in that a shared–use path also has perceived lanes of travel. The model estimates the number of passings and meetings by a test bicyclist traveling at the mean speed of bicyclists on the trail. "Meetings" refer to users traveling in the opposite direction of the test bicyclist, and "passings" occur when the test bicyclist overtakes users traveling in the same direction. The Botma procedure, as adopted in the 2000 HCM, compiles the numbers of bicyclists and pedestrians who are met and who are passed. The LOS of bicycles is determined by adding the number of meetings estimated to twice the number of passings estimated and comparing this number of weighted events to an LOS scale. For an LOS scale, Rouphail, et al., recommended the use of the A through F scale, which is familiar from other chapters of the HCM, with essentially arbitrary boundaries between levels.(3)

Limitations of Current LOS Method

As described above, Botma's procedure, which bases LOS on the estimated number of meetings and passings for bicyclists, is an attractive framework. There can be little debate that, in general, paths where bicyclists incur more meetings and passings should be less desirable than trails with fewer meetings and passings. However, the LOS procedure in the 2000 HCM has a number of serious limitations that make it difficult for designers to use in resolving their path design dilemmas. These limitations include:

  • The procedure needs to be calibrated and validated for U.S. conditions. As detailed in chapter 2, Botma's equations are based on sound theory and they are based, in part, on field data from The Netherlands.(5) However, they have never been compared to U.S. field data. U.S. paths are typically wider than European paths. U.S. bicyclists are generally not as experienced. U.S. bicyclists tend to ride more often for recreation and less often for commuting. And U.S. bicycles are different from European bicycles. Among the parts of the model that need to be calibrated is the relative weighting of passings to meetings.

  • The procedure does not account for "passive passings." This is an event when the test bicyclist is passed by a faster path user. Passive passings are probably undesirable from the test bicyclist's perspective and should be considered in an LOS procedure.

  • The procedure assumes that path users do not impede each other's movements (i.e., that there is always adequate room for the test bicyclist to pass with no change in speed or lateral positioning). This is true only if: (1) the path is wide enough, and/or (2) there is no opposing traffic during the passing maneuver. If passing is restricted, there will be a number of "delayed overtakings." The significance of this limitation cannot be overstated. When passing and meeting become restricted, the procedure cannot predict that the LOS will worsen, because the number of events actually decreases. The procedure also assumes that bicyclists always want to pass any encountered bicyclists or pedestrians who are going at a slower rate. However, if the speed difference is small, or if the test bicyclist is near the end of his or her time on the path, this will not be true.

  • The current LOS procedure for shared–use paths accounts for pedestrians and bicycles only. However, in his original model, Botma simulated other path users, including mopeds and tandem bicycles.(5) Shared–use paths in the United States currently accommodate large numbers of joggers, inline skaters, skateboarders, and other types of users. The addition of other path users can be represented analytically in one of two ways. If a path user group appears to have a similar mean speed to another group, then such groups can be combined into one larger group that has a common mean and standard deviation. However, if a group is quite different from the others, then all events associated with this group must be estimated using separate equations.

  • The current procedure is based on single values of mean bicycle and pedestrian speed. Designers in areas where bicycles and pedestrians may travel faster or slower should have the ability to incorporate that information into their LOS estimates.

  • The 2000 HCM method is limited to the analysis of two–lane and three–lane paths. Furthermore, two–lane paths are specified as 2.44 m (8 ft) wide, and three–lane paths are specified as 3.05 m (10 ft) wide. Designers considering other widths and numbers of lanes have no current guidance.

  • A stronger basis for the LOS criteria is clearly needed. While most of the LOS criteria in the 2000 HCM were set based on the expert opinion of the members of the Highway Capacity and Quality of Service Committee, there is recent research in some areas that bases the criteria on user surveys. In the pedestrian and bicycle arena, Harkey, et al., and Landis, et al., both members of this research team, have developed LOS criteria for on–street bicycle paths that are validated against user perceptions of the quality of service.(67) A set of LOS criteria that is well grounded in user perceptions would be more credible than a set based on expert opinion alone.

  • There is a great need to effectively convey the procedure and criteria to shared–use path designers and operators. The HCM should certainly remain as one way to convey the procedure and criteria. However, the HCM is not a prominent document among shared–use path designers and operators. Also, the next version of the HCM may not be issued for many years. We need to convey any new procedure to the users in an effect manner and sooner.

With all of these limitations on the current LOS procedure, the need is clear for a substantial research effort to refine the method and to provide designers with a new procedure.



The overall project objective was the production of a tool that professionals can use to evaluate the operational effectiveness of a shared–use path, given a traffic forecast or observation at an existing path along with some geometric parameters. The project adopted Botma's method as the basic framework for the LOS procedure.(5) In particular, the objective was to produce a tool that would overcome the major limitations in the current LOS procedure noted above. It was desirable that the procedure emerging from this project would:

  • Be calibrated and validated.
  • Be based on U.S. data.
  • Have LOS criteria based on user input for a typical mix of trip purposes.
  • Include more modes.
  • Include the ability to change key parameters such as mean speeds.
  • Account for delayed passing.
  • Analyze the full range of existing and possible path widths.
  • Be in a form ready for use by path designers.



The four major activities needed to achieve the project objective described were:

  1. Development of the additional theoretical framework necessary to overcome the limitations of the existing procedure noted above.

  2. Collection of field data on path operations to calibrate and validate the theoretical equations for U.S. conditions.

  3. Collection of path user perception data to establish LOS criteria.

  4. Development of an LOS estimation tool that professionals working with shared–use paths could use, and a plan to distribute that tool.

Development of Theory

To achieve its objective, the project team had to develop the theory of traffic flow on shared–use paths in two important ways. First, the team had to determine a way to calculate the number of passive passings that occurred on a typical path. As noted above, a passive passing is an occasion when the test bicyclist is passed by a faster path user. Because bicyclists are typically the fastest users on a path, the number of passive passings is probably small in most cases; however, it should contribute to an LOS estimate. Passive passings were not used in the 2000 HCM procedure.

Furthermore, the team had to find a way to calculate the number of delayed passings. These are times when the test bicyclist would arrive behind a slower path user and not be able to pass because of the lack of an adequate–sized gap in the next lane to the left (oncoming or same direction). Obviously, delayed passings are undesirable for bicyclists since they would have to slow down and then expend energy accelerating when an adequate gap appears. Delayed passings are also critical because they are so closely related to path width. Prior to this project, there were some delayed passing calculations in the literature related to two–lane highway operation and similar facilities; however, nothing in the literature related to shared–path operation.

Operational Data Collection

The objective of the operational data collection portion of this project was to collect the field data needed to calibrate and validate the LOS model for shared–use paths. To calibrate and validate an LOS model, the main variables that needed to be collected were meetings and desired and actual passings by path users. Other data that need to be collected are the mean speed and speed range of the different user groups. In addition, trail characteristics must be recorded at each site. To ensure later flexibility, it was desirable that scenes on paths of interest be recorded from different perspectives so that additional data could be obtained by viewing videotapes if needed.

The project proposal identified three methods of data collection: (1) a one–camera method, (2) a two–camera method, and (3) a moving–bicycle method. The one–camera method placed a camera at an elevated position where it could record scenes on a long path segment. The two–camera method recorded when path users entered and exited a path segment of interest, inferring meetings and passings on that segment. The moving–bicycle method, by contrast, collected meetings and passings from the perspective of a test bicyclist using a camera mounted on the bicyclist's helmet.

After careful consideration of all of the pros and cons for all three methods, the team chose to use the moving–bicycle method. Vantage points for the one–camera method would be rare (tall buildings and hills with unobstructed views of qualifying shared–use paths are not common in the United States). The two–camera method would not be able to identify the difference between actual passings and desired passings because only path users would know whether they wanted to pass and were unable to do so and why. For example, a bicyclist may not have been able to pass because of inadequate path width or congestion. The moving–bicycle method can collect the needed data without these problems. The moving–bicycle method can be supplemented with a stationary camera on the side of the path. It can also be used to record path user volumes and the characteristics of different user groups, such as mean speed. Consequently, the moving–bicycle method, supplemented by a stationary camera, was the primary operational data collection method.

Perceptual Data Collection

A major part of this effort was to help set the LOS criteria by collecting data on user perceptions of multi–use trail design and operations. From a user perception standpoint, the intent of the present study was to quantify the effect of selected operational trail parameters on bicyclist and pedestrian judgments of the perceived adequacy of the trail facility. It is recognized that user responses will differ, depending on the individuals' own reasons for using the trail (e.g., whether they were seeking a casual and relaxed activity or a rigorous individual workout unencumbered by users with more relaxed intentions). It was beyond the scope of this study to collect data on user perceptions as a function of users' individual intentions or needs. However, an effort was made to obtain the opinions of a variety of users.

The research team believed that it was possible to define the LOS for a trail in operational terms, independent of the factors governing the capacity of the trail. For example, a two–lane trail will obviously have less capacity than a four–lane trail; however, both, under different demand conditions, may be described as operating at the same LOS. In the present study, LOS is assumed to vary as a function of operational trail conditions that can be specified largely in terms of meeting and passing events. Depending upon the capacity of a trail and its particular level of use, each trail can be described in terms of the frequency of these meeting and passing events. If it could be shown that users' judgments of the adequacy of a trail vary as a function of such events, it would be possible to predict user response to trail conditions and designs beyond the limited set of paths addressed by the study.

A Usable Procedure

As noted above, an important element of this project was that the procedure developed was usable by trail design professionals and that it would be distributed in a manner that would reach them. The research team included trail design professionals who carefully crafted the products for their colleagues. In addition, the researchers developed products that could be adopted in future versions of the HCM, as well as distributed in other ways. A section later in this chapter describes the research products in more detail.



The scope of the project and, therefore, the products emerging from the project, was limited in several important ways. First, the project was limited to selected nonmotorized travel modes. We set out to expand the current procedure to those nonmotorized modes that are common on typical U.S. shared–use paths. In the end, we collected data on adult bicyclists, child bicyclists, walking pedestrians, running pedestrians, and inline skaters, and included these in our LOS method. Other modes of travel seen occasionally on shared–use paths, such as roller skaters, scooters, wheelchairs, Segways, and tandem bicycles, were not included because we did not see enough of them during our data collection for inclusion. Riders on horseback and snowmobiles are examples of other occasional path users that were outside the scope of this effort because they did not use the paths of interest in large numbers year–round. The LOS estimation procedure could be expanded to include any of these modes, or any other mode, if the analyst possessed some basic data about the mode, such as mean speed.

The scope of the project was also limited to off–street, paved paths. Although the methodology developed could apply to paths used exclusively by bicyclists and to one–way paths, the bulk of the attention in this research was centered on two–way paths serving pedestrians, bicyclists, and other users because they are the vast majority of the off–street paths in the United States. Since most paths with gravel, dirt, wood chips, or other loose material on the surface do not attract much bicycle volume, project data collection and analysis were limited to paths that were paved or had hard surfaces. A designer who is working on a path that has a hard–packed gravel or granular stone surface on which bicyclists operate in a very similar manner to paved paths may be able to apply the methodology we developed for that path with minimal additional error.

Also, the project scope was limited in that the LOS produced was from the bicyclist's point of view. The researchers collected some perception data from the pedestrian's point of view, but not enough to establish their own LOS scale. Chapter 9 will recommend future research targeted at estimating path LOS from the points of view of pedestrians, skaters, and others.

Furthermore, the project scope was limited to the analysis of trail segments at least 0.40 kilometers (km) (0.25 miles (mi)) long, uninterrupted by stop signs, signals, important intersections, or other similar features. Analysts will need other ways to find the LOS at these points.

Finally, the project results were not intended for forecasting the number of future users of a path. While there may be some overlap between operational/design and forecasting methods, the premise behind this effort is that user volumes are an input rather than an output. The intent of this project was to answer questions regarding how wide the path should be to satisfy current or future demand, rather than to estimate how many users would be attracted to a path of a certain design.



Description of the Products

The three final products of this study are: (1) this report, (2) Share–Use Path Level of Service Calculator: A User's Guide (User's Guide) (Publication No. FHWA–HRT–05–138), and (3) Evaluation of Safety, Design, and Operation of Shared–Use Paths, a TechBrief (Publication No. FHWA–HRT–05–139). These products will be distributed primarily via the U.S. Department of Transportation (USDOT) Pedestrian and Bicycle Information Center (PBIC) Web site. The User's Guide provides detailed, step–by–step instructions on how to use the LOS procedure and spreadsheet calculation tool, which can be downloaded from the Turner–Fairbank Highway Research Center Web site at www.tfhrc.gov. The User's Guide and TechBrief can also be downloaded from the Web site.

The widespread use and application of the LOS methodology is ultimately dependent upon how easy it is to use, whether it is considered applicable to trail design scenarios, and whether trail designers are able to gather the data needed to use the model. At this time, there are two main applications of the model: (1) to determine the appropriate width of a new trail, and (2) to determine how much width to add to an existing trail to accommodate current or projected levels of use. Determining whether to separate modes or directions of travel is also emerging as a key application.

The availability of data and its ease of collection are often key components in the success of an LOS model. We tried to ensure that the data items collected for the model would be relatively easy for a trail designer to obtain. We also recommended default values for most of the needed inputs. In addition, the User's Guide describes how to effectively collect data for use in the model.

Our idea for the LOS calculator was that it should be some type of spreadsheet application or self–executing graphical user interface software. The team was inspired by the League of Illinois Bicyclists, which developed online graphical user interface software that calculates a bicycle LOS for a roadway using the Bicycle Compatibility Index and the Bicycle LOS model (see http://www.bikelib.org/roads/blos/losform.cfm). (67)

The interface is easy to use, is accessible directly from the League's Web site, and suggests default values if an analyst does not have all of the necessary data. A user can simply click on the calculate button, and an LOS result for each model is displayed. We attempted to create a calculator for our shared–use path LOS model that would be made available in a similar format and would allow users to easily calculate an LOS for a shared–use path.

Intended Users

Unlike the roadway environment, which is almost exclusively the domain of civil engineers, shared–use paths are designed by a wide variety of practitioners. Some of the most creative and unique trails in the country are the direct result of the diverse skills of these designers. Since these professionals look to a variety of different sources for design guidance, establishing national guidelines is difficult.

We identified three main target audiences for the marketing of our shared–use path LOS model: (1) transportation professionals, (2) trail designers/coordinators, and (3) pedestrian, bicycle, and trail advocates and organizations:

  • Transportation Professionals. These individuals are engineers, planners, or designers. They work in planning, engineering, and public works departments at all levels of government. They may also work for consulting firms or at research institutes. These individuals often rely on roadway design manuals such as the AASHTO Guide for the Development of Bicycle Facilities, the Highway Capacity Manual, the AASHTO Policy on Geometric Design of Highways and Streets (the Green Book), and the Manual on Uniform Traffic Control Devices (MUTCD), along with other State and local roadway design manuals. (See references 1, 4, 8, and 9.) Transportation professionals are more likely to possess a technical background in standard LOS applications, roadway cross sections, and in the design of roadways and/or bicycle and pedestrian facilities.

  • Trail Designers/Coordinators. These individuals are planners, landscape architects, or other professionals who are involved in the design, development, and maintenance of trails; however, they may not have a technical background in transportation planning. They work in planning, parks and recreation, environmental protection, or greenway and trail departments at all levels of government, and they also may be consultants hired by governments to design shared–use paths. They often rely on park and recreation design manuals, information about the design of trails obtained from various clearinghouses, and past professional experience. This group is important to reach because they often make decisions about the design, location, and development of shared–use paths.

  • Pedestrian, Bicycle, and Trail Advocates and Organizations. These groups often play an essential role in making shared–use paths a reality. Organizations such as Rails–to–Trails Conservancy and East Coast Greenway Alliance often provide technical guidance and/or serve as clearinghouses for innovative design approaches. Local trail alliance/advocacy groups are often influential, and many times they make major decisions in the development of shared–use paths. It is important that the shared–use path LOS model address their concerns and that it be embraced as a useful tool by these groups as well. These groups frequently set a vision for trails in a community, provide input into the kinds of trails to be created, and provide coordination among the many players who will develop, own, and manage the trail.

Through this report, the User's Guide, and the TechBrief, we tried to reach all three of these groups. It should be noted, however, that the primary intent of this report is to provide technical details with regard to our methods and data. Unlike the User's Guide and the TechBrief, this report is not intended for wide distribution.



This report includes eight chapters in addition to the introductory chapter. Chapter 2 is a review of the literature pertaining to LOS estimation for shared–use paths. Chapter 3 describes the development of the theoretical background that we needed for the procedure. Chapter 4 discusses the methods we used to collect the field data on path operations. Chapter 5 shows how we used the field data to calibrate and validate our LOS model. Chapter 6 describes how we collected data on user perception of shared–use paths having various geometric and operational characteristics. Chapter 7 shows how we analyzed the perception data in order to develop the LOS criteria. Chapter 8 presents the highlights from the LOS procedure. (A much more comprehensive guide to the procedure, written for the audiences we described above, is available in the User's Guide.) Finally, chapter 9 provides a summary of the project and our recommendations for future research to improve technical capabilities in this area. At the end of the report are a set of appendixes and a complete list of references.



Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101