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-05-138
Date: July 2006

Shared-Use Path Level of Service Calculator

A User's Guide

PDF Version (1298 KB)

PDF files can be viewed with the Acrobat® Reader®





During the design of every shared-use path, someone eventually asks, "How wide should this 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?"

At the present time, conventional design manuals provide little guidance on these issues. 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 3.0 m (10.0 feet).... Under certain conditions it may be necessary or desirable to increase the width of a shared-use path to 3.6 m (12.0 feet) or even 4.2 m (14.0 feet), due to substantial use by bicycles, joggers, skaters and pedestrians."(2) No further guidance is given to determine what specific levels of use–or mixture of uses–warrants a wider pathway or a separation of users. The purpose of this research is to fill this information gap and to give planners, designers, and managers the necessary tools to make more informed decisions regarding trail width and design.

Previous research efforts have laid the groundwork for the study of this problem. In order to provide some level of guidance in the year 2000 edition of the FHWA Highway Capacity Manual (HCM)(3), Rouphail, et al., recommended an analytical procedure to determine LOS for bicyclists on shared off-street paths.(4) Rouphail, et al., adapted a procedure originally developed by Hein Botma(5) in the Netherlands.(4,5)

Botma is considered an important pioneer in the development of early measures of user density (crowding), flow characteristics, and hindrance for bicyclists using shared-use paths. His procedure bases LOS on two factors: first, the number of trail users passed in the opposing direction (meetings), and second, the number of slower trail users that are traveling in the same direction that a bicyclist must overtake (active passes). A useful starting point, Botma's framework had a variety of limitations for application in the United States.(5)

The Botma method addressed only two trail user modes–bicycles and pedestrians. Runners, in-line skaters, and other common path users were not incorporated. The significant presence of these path users on U.S. trails suggested that a tool developed for application in the United States should attempt to address them.

Next, the LOS scores generated by the Botma method were developed based on his experience of shared-use paths in the Netherlands and were not calibrated to the opinions of actual trail users.(5) Harkey, et al., (1998) and Landis, et al., (1997) have developed LOS criteria for on-street bicycling conditions, which are validated against user perceptions of the quality of service.(6,7) The success of their work has shown that when quantitative measures are statistically calibrated with user perceptions, the resulting model has greater acceptance with stakeholders at all levels–technical staff, elected officials, facility users, and the general public.

Finally, Botma's work did not include investigation of any other factors to determine if they might have a significant influence on trail users' perceived LOS.(5) Factors such as delay, trail-striping patterns, the presence of usable shoulders, sight distance, or lateral clearance all might be important. Identification, analysis, and incorporation of other criteria that may contribute to actual users' opinion of LOS will ensure the development of a more effective predictive tool.



The purpose of this study was to gather and analyze background data necessary to develop a model that professionals can use to evaluate the operational effectiveness of a shared-use path, create and test that model, and develop a user-friendly interface and guide for its use.



To accomplish this purpose, a three-step approach was developed:

  1. Review prior research and develop a sound theoretical approach to solving the problem.
  2. Gather the necessary operational and user perception data.
  3. Analyze the data and develop and test a model that can be used by practitioners as a planning and design tool.


Step 1: Develop Theoretical Approach

Step one was accomplished through the completion of a literature review and the development of a technical dissertation that described the theoretical approach. The results of these interim products are summarized separately from this User's Guide, in the project's Final Report.(8)


Step 2: Data Collection

Step two involved gathering and analyzing two types of data, operational data and user perceptions. Operational data collection involved gathering volume, speed, and mode split data from test locations along 15 trails throughout the United States. It also included gathering sufficient information about the characteristics of each study trail to create a thorough path profile and also testing other factors that might prove influential to bicyclists' LOS. These data included path width, surface type, path setting, sight distance, lateral clearance, presence of a shoulder or adjacent unpaved treadway, and other factors. See appendix A for profiles of the 15 study trails.

User perception data were gathered by surveying over 100 typical trail users regarding their perceptions of four aspects of shared-use path operations:

  • Lateral spacing.
  • Longitudinal spacing.
  • Ability to pass.
  • Overall perception of comfort and freedom to maneuver.

Designing this aspect of the research posed a challenge: how to have a group of approximately 100 survey participants experience a wide variety of trail conditions. Because cost and logistics made the physical movement of groups of respondents to multiple trail survey sites impossible, it was decided to bring the study trail sites to the survey participants through the use of videotaped trail experiences.

A total of 105 survey participants were gathered in two metropolitan areas–Raleigh-Durham-Chapel Hill, NC, and Washington, DC. Each group of participants viewed 36 one-minute video clips of a test bicyclist traveling along a section of one of the study trails.[3] By viewing exactly the same video clips, two geographically distinct survey groups vicariously experienced the same variety of trails and trail conditions.

The video clips were created using a helmet-mounted video camera. A member of the research team (the test bicyclist) bicycled down a 0.8-km (0.5-mi) segment of each of the study trails while the camera taped the view of the trail ahead. Approximately 60 three-minute video clips were created for each of the study trails.[4] The 36 one-minute video clips used for the survey were selected from this footage.

Survey participants viewing the video clips could see the path ahead, the landscape on each side of the trail, the oncoming trail traffic, and slower moving traffic in the test bicyclist's lane. At times, the video clips included a tape of passing maneuvers, when the test bicyclist overtook slower moving traffic. The test bicyclist tried to remain at a constant speed for the duration of each video clip.

The participants in the survey were asked to score each video clip for the four aspects noted previously: lateral spacing, longitudinal spacing, ability to pass, and overall perception of comfort and freedom to maneuver.


Step 3: Data Analysis/Development of the Model

The raw operational data were analyzed for three purposes: to develop volume and mode split profiles for each study trail; to develop a variety of constants and other factors for use in the LOS model; and to validate the theoretical methods developed to predict meetings and passings. The constants gathered included average and standard deviation speeds for all trail-user types, a peak-hour factor, a propensity-to-pass factor, and passing-distance intervals.

The user perception survey results were expected to correlate with measurable events on each trail. Each trail was analyzed for the following four measurable events:

  • Meetings: the number of trail users (by user type) that passed the test bicyclists going in the opposite direction.
  • Active passes: the number of users traveling in the same direction (by user type) that were passed by the test bicyclists.
  • Passive passes: the number of times the test bicyclist was passed by trail users traveling in the same direction.
  • Delayed passes: the number of times that the test bicyclist needed to pass in order to maintain speed but was blocked by other users traveling in either direction.

A full discussion of model development and creation of the Shared-Use Path LOS Calculator is presented in chapter 3.



The most important findings of this study can be organized into two groups: characteristics of study trails and perception survey results.


Characteristics of Study Trails

Typical 2-way trail user volumes were found to vary considerably from as low as 43 per hour on the W&OD Trail in Northern Virginia to 2,316 per hour on the North Beach Lakefront Path in Chicago, IL. [5] Per-hour volume data were created by averaging the counts from approximately 60 three-minute trials taken over the course of a peak usage day (from about 10 a.m. to 5 p.m.), typically Saturday or Sunday. Where commuter traffic was significant, some data were collected on weekday mornings. See table 1 for a summary of average per-hour volumes by study trail.


Table 1. Variations in trail user volume.

Path name Location Average two-way volume (per hour) Trail width (ft) Successful data collection trials (3 minutes each)
W&OD Trail Arlington, VA 44 10.0 4
Honeymoon Island Trail Dunedin, FL 110 12.0 48
Pinellas Trail Pinellas County, FL 120 15.0 57
Grant's Trail St. Louis County, MO 122 12.0 30
Capital Crescent Trail Washington, DC 159 10.0 9
Lake Johnson Trail Raleigh, NC 205 8.0 58
White Creek Trail Dallas, TX 216 8.0 60
White Rock Lake Trail Dallas, TX 252 14.0 60
Forest Park Trail St. Louis, MO 299 10.0 57
Sammamish River Trail Redmond, WA 418 10.0 58
Charles River Bike Path Boston, MA 438 8.0 60
Minuteman Bikeway Arlington, MA 442 12.0 60
South Bay Trail Santa Monica, CA 616 14.0 60
Mill Valley–
Sausalito Pathway
Marin County, CA 641 9.5 60
Lakefront Trail Chicago, IL 2320 20.0 90

1 ft = 0.30 m


User mix also varied considerably on the 15 study trails. Users on five trails were observed to comprise more than 70 percent adult bicyclists. On six trails, more than 35 percent of users traveled on foot (pedestrians and runners). Observation of user types found few users outside of the five basic user groups presented in table 2. Wheelchair users and push scooters were observed on some trails, but not in sufficient quantity to develop a statistically valid average speed profile.


Table 2. Variations in trail user mix (mode split).


Average percent of volume by trail user type

(mode split)

Path name Location Trail width (ft) Adult bicyclists Pedes-trians Runners Skaters Child bicyclists
Pinellas Trail Pinellas County, FL 15.0 81.4 4.6 2.3 11.6 0.0
Sammamish River Trail Redmond, WA 10.0 78.9 3.4 3.4 6.0 8.4
W&OD Trail Arlington, VA 10.0 73.7 5.3 15.8 5.3 0.0
Charles River Bike Path Boston, MA 8.0 72.3 8.2 3.8 14.7 1.1
White Rock Lake Trail Dallas, TX 14.0 71.6 13.6 8.0 3.4 3.4
White Creek Trail Dallas, TX 8.0 64.8 9.9 6.6 14.3 4.4
Mill Valley–Sausalito Pathway Marin County, CA 9.5 62.8 7.8 27.8 0.0 1.7
Grant's Trail St. Louis County, MO 12.0 59.2 16.3 4.1 10.2 10.2
Capital Crescent Trail Washington, DC 10.0 55.9 17.0 18.6 3.4 5.1
Minuteman Bikeway Arlington, MA 12.0 51.9 6.2 15.6 18.1 8.1
Lakefront Trail Chicago, IL 20.0 48.8 20.5 17.7 12.3 0.7
South Bay Trail Santa Monica, CA 14.0 40.3 17.4 12.5 25.0 4.9
Forest Park Trail St. Louis, MO 10.0 33.0 24.4 27.8 13.9 0.9
Honeymoon Island Trail Dunedin, FL 12.0 22.9 54.2 12.5 8.3 2.1
Lake Johnson Trail Raleigh, NC 8.0 14.1 63.3 21.9 0.0 0.8

1 ft = 0.30 m


Based on the width, centerline, volume, and mode split data collected on the study trails, an average trail profile was developed. To ensure that the outliers in the data set did not skew the average, the high- and low-volume trails (Lakefront and W&OD, respectively) were not used to calculate the data profile for the average trail in table 3. Mode split shares were rounded to the nearest five.


Table 3. Data profile for the average trail.

  Width Centerline One-way volume
per hour
Adult bicycles Pedestrians Runners Skaters Child bicycles
Average trail 11 ft Yes 105 55% 20% 10% 10% 5%

1 ft = 0.30 m


This research also yielded a large volume of data on average speeds for different types of trail users. Table 4 shows the average user speed for each mode and the typical range of variation (standard deviation).


Table 4. Average speed by mode.

Trail user type (mode) Average speed (mi/h) Standard deviation (mi/h)
Adult bicyclists 12.8 3.4
In-line skaters 10.1 2.7
Child bicyclists 7.9 1.9
Runners 6.5 1.2
Pedestrians 3.4 0.6

1 mi/h = 1.6 km/h


Perception Survey Results

As explained above, a survey of 105 trail users (primarily bicyclists) was used to determine what factors bicyclists found to be significant in their evaluation of comfort and freedom to maneuver on shared-use paths. Using standard statistical methods, a wide variety of factors were tested to determine their overall influence on survey responses. The following is a summary of these findings.

The primary factors found to affect bicyclists' perceived LOS were:

  • Path width.
  • Active passes (frequency of encountering and passing other users in the same direction).
  • Meetings (frequency of encountering other users in the opposite direction).
  • The presence of a striped centerline.

The frequency of active passes and meetings were determined by user mix (mode split) and the overall user volume on the trail.

As a result of the survey, bicyclists' LOS was shown to be most affected by sharing a trail with slower users. Pedestrians had the greatest negative impact, because they had the slowest average speed. For example, a bicyclist traveling at 19.3 kilometers per hour (km/h) (12.0 miles per hour (mi/h)) faced an impediment when encountering a pair of pedestrians traveling at 4.8 km/h (3.0 mi/h) or a runner traveling at 9.76 km/h (6.0 mi/h). During these encounters, the bicyclists executed passing maneuvers to maintain speed.

Encountering significant numbers of slow moving users in the same direction of travel increased the need to make passing maneuvers. Encountering significant numbers of slow moving users traveling in the opposite direction tended to block the space needed to make passing maneuvers. Width played a factor in determining how much space was available to make passing maneuvers, and the presence of a striped centerline was also found to affect the bicyclist's sense of freedom to maneuver.

In summary, bicyclists' LOS decreased when:

  • The need to pass other users increases.
  • The amount of space available to make a passing move decreases.
  • The probability that a passing opportunity will be blocked by other users increases.

Factors found to have little or no effect on the bicyclist's operational comfort include the following:

  • Trail setting.
  • Lateral clearance.
  • Sight distance.
  • Presence of a shoulder.
  • Presence of horizontal curves.

Each of the factors listed above was evaluated only within the range of variation extant among the 15 study trails (see appendix A for details). Within those ranges of variance, these factors were not found to be statistically correlated to the LOS ratings given by the participants in the user perception survey.



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