1. Questions
Bicycle parking is a key component of effective cycling systems and urban planning (Egan et al. 2023; Honey-Rosés et al. 2023). However, many studies focus only on whether bicycle parking was present or absent without accounting for the quality of parking (Heinen and Buehler 2019). Inadequate parking can leave bicycles exposed to theft and damage, and can deter people from cycling (Chen et al. 2018; Cohen et al. 2024). To address this, we recorded rack type, capacity, and condition of bicycle parking in addition to location. Studies also commonly focus on transit stations or workplaces, while fewer studies cover bicycle parking across entire municipalities (Ito et al. 2023; Honey-Rosés et al. 2023; Heinen and Buehler 2019). This study examines bicycle parking across all town-owned property in Lexington, Massachusetts. Lexington contains established bicycle infrastructure including the Minuteman Bikeway that connects to Alewife station in Cambridge, and the town supports bike travel through the Safe Routes to School Program (Minuteman Bikeway, n.d.; Town of Lexington, n.d.). Bicycle parking has been added to town property over many years, but no inventory existed of what was installed, where, or in what condition.
We ask: How do the capacity, condition, rack type, and location of bicycle parking vary across town-owned sites in Lexington, Massachusetts?
2. Methods
A field survey of bicycle parking on Town of Lexington property was conducted from September 8, 2025, to July 2, 2026. Every bicycle rack was located and recorded using ArcGIS Survey123, a mobile application that logged the position of each rack. Survey data was compiled and analyzed in Excel, and maps were created in ArcGIS Pro.
The survey covered schools, parks, conservation land, civic buildings, the Minuteman Bikeway, town trails, and the public right-of-way along streets. Racks on private property, including businesses, apartment buildings, and private schools, were not surveyed. Town-owned parcels were identified from the Massachusetts Property Tax Parcels data layer (MassGIS 2025). Every parcel was visited, and those with no rack were recorded as such.
For each rack, the rack type, capacity, and condition were recorded. A rack was defined as one freestanding unit, or a group of units of the same type in the same spot. Where two rack types existed in the same spot, each was recorded separately.
As shown in Figure 1, rack type was divided into six categories: inverted-U, triangle/schoolyard, grid, wave, decorative/custom, and bollard/post-and-ring. Capacity was counted as two bikes for each element of a rack and one bike per slot on grid racks. Note S3 describes the rule for each type. Total capacity, rounded mean capacity, and number of racks were reported by type.
Racks were rated on four criteria: whether the rack was bent, whether the metal was sound, whether it could be reached without obstruction, and whether a bike could be locked to it. Each criterion was rated as good, fair, or poor. The overall condition of the rack was the lowest of the four ratings. Note S4 describes how the ratings were determined. The number of racks in each condition was reported.
The racks were grouped by the nearest place they served. Racks that did not serve a specific place were assigned to the street or trail where they were located. Racks at multi-rack sites were numbered north to south. Racks and capacity at multi-rack sites are reported in Table S1.
Each site was assigned one of five settings: schools, parks and open space, civic buildings, streets, and bikeway and trails. The number of sites, racks, and total spaces were reported by setting.
Capacity per 100 students was calculated for each of the nine K-12 public schools in Lexington, Massachusetts. Enrollment data for 2025–26 was obtained from the Massachusetts Department of Elementary and Secondary Education (2025).
The inventory is provided in Table S2 and a supplementary CSV, and mapped on the town website (Town of Lexington 2026).
3. Findings
Figure 2 displays the location of bike racks and their capacity.
Capacity was concentrated within the area bounded by Route 128 and Route 2, and clustered around Lexington Center. Most bike racks were close to shared-use paths or the Minuteman Bikeway. However, several high-capacity racks were in the town periphery at schools without nearby shared-use paths.
Table 1 shows the number of racks and total spaces by type.
Inverted-U racks made up nearly half of the 99 racks but held only 38 percent of the spaces. Grid and schoolyard racks were 38 of the 99 but held 49 percent of the spaces.
Table 2 shows bike parking by setting.
Capacity was concentrated at a few sites. The ten largest of the 41 sites held 71 percent of spaces, while the 24 sites with a single rack held 233 spaces, or 21 percent. More sites in a setting did not correspond to more parking. Parks and open space was the most common setting, with 19 of the 41 sites, but most had only one or two racks, and together they held 13 percent of spaces. There were fewer school sites, yet each usually had several large racks. Schools averaged 74 spaces per site, while parks and open space averaged 8. Most of the highest-capacity racks were on school grounds.
Figure 3 maps bicycle parking by setting.
Schools held nearly two-thirds of bike parking. Table 3 compares each school’s capacity with its enrollment.
Lexington High School had more bike parking spaces than any other site and the second fewest per 100 students. The three schools with the most parking per 100 students were all elementary schools.
Of the 99 racks, 88 were in good condition, 7 were fair, and 4 were poor. The four poor racks were all grid or schoolyard racks, and they held 72 of the 1,134 spaces.
There is now a record of the location, type, capacity, and condition of bicycle parking on municipal property. However, the survey did not measure how much of the parking supply is used. The approach could serve as a framework for other towns and future surveys and analysis of bicycle parking.
Acknowledgments
Thank you to Ben Alperovich, Rhys Behling, Jayant Krishnan, and Molly Eriksson for helping collect the survey data and to Joshua King for his support and advice throughout the research. Anthropic’s Claude was used to edit the manuscript for grammar and clarity. The author takes full responsibility for the content of the publication. This research received no funding, and the author declares no conflict of interest.



