Classification Categories

The American Walkability Atlas classifies neighborhoods into seven walkability categories based on National Walkability Index scores. The categories provide additional detail beyond the EPA’s four broad classifications while maintaining the same underlying National Walkability Index methodology. They describe how the built environment supports walking through the combined effects of street connectivity, transit accessibility, employment diversity, and the mix of jobs and households. The National Walkability Index itself is calculated using these four variables developed by the U.S. Environmental Protection Agency. Research has consistently shown that neighborhoods combining connected street networks with mixed land uses and accessible destinations encourage greater walking, improve accessibility, and support healthier travel behavior (Cervero & Kockelman, 1997; Ewing & Cervero, 2010; Frank et al., 2005)

• Very High Walkability (17.143–20.000): Compact, mixed-use neighborhoods with highly connected street networks, abundant destinations, and excellent transit access where walking is a practical choice for many daily trips. These neighborhoods typically contain downtown districts, historic urban centers, or major mixed-use activity centers.
• High Walkability (14.286–17.143): Well-connected neighborhoods with diverse land uses, strong street connectivity, and convenient access to employment, shopping, parks, schools, and transit. Walking is practical for many everyday activities, although some trips may still require driving or transit.
• Above Average Walkability (11.429–14.286): Balanced neighborhoods where walking and driving both play important roles. Residents generally have reasonable access to everyday destinations through moderately connected street networks and a mix of residential and commercial development, although destinations may be more dispersed than in highly walkable areas.
• Moderate Walkability (8.571–11.429): Neighborhoods where walking is practical for nearby destinations but automobile travel becomes increasingly necessary for longer trips. These areas often combine older walkable sections with newer suburban development, lower intersection density, or reduced transit accessibility.
• Below Average Walkability (5.714–8.571): Predominantly suburban neighborhoods characterized by separated land uses, limited destination diversity, lower street connectivity, and greater dependence on automobiles. Walking is generally limited to recreation or short neighborhood trips.
• Low Walkability (2.857–5.714): Low-density neighborhoods with disconnected street networks, limited nearby destinations, and minimal transit access. Most daily activities require driving because employment, shopping, and services are located relatively far from residential areas.
• Very Low Walkability (1.000–2.857): Rural areas, large undeveloped landscapes, airport districts, industrial areas, parks, or highly dispersed suburban neighborhoods where the built environment provides very little support for walking. These locations typically have very low street connectivity, few destinations, and limited pedestrian infrastructure.

Planning Implications

The National Walkability Index provides planners with a valuable screening tool for evaluating how neighborhood design supports walking and access to daily destinations. Because the index combines street connectivity, transit accessibility, employment diversity, and the balance between jobs and housing, it helps identify neighborhoods where the built environment already supports walking and those where targeted improvements may be needed. The EPA developed the index specifically to support land-use planning, transportation planning, public health initiatives, and scenario analysis by providing a consistent nationwide measure of neighborhood walkability.

Neighborhoods classified as Very High or High Walkability typically represent established urban centers, historic neighborhoods, transit-oriented developments, or mixed-use districts where planning efforts should focus on preserving pedestrian infrastructure, maintaining transit service, encouraging infill development, and protecting neighborhood accessibility. These areas often serve as models for future walkable development while also requiring strategies to address housing affordability and displacement pressures associated with highly desirable walkable environments.

Neighborhoods within the Above Average and Moderate Walkability categories often present the greatest opportunity for incremental improvements. Investments such as new pedestrian crossings, sidewalk gap closures, bicycle infrastructure, transit enhancements, mixed-use redevelopment, and improved street connectivity can substantially increase walkability without requiring major redevelopment. Research by Cervero and Kockelman (1997) demonstrates that improving density, land-use diversity, and street design can significantly influence travel behavior, while Frank et al. (2005) found that neighborhoods combining connected streets with mixed land uses encourage higher levels of walking and physical activity.

For neighborhoods classified as Below Average, Low, or Very Low Walkability, long-term planning strategies may focus on improving connectivity, increasing destination diversity, expanding transit service, encouraging mixed-use development, and reducing automobile dependence where appropriate. However, planners should recognize that some low-scoring neighborhoods consist primarily of airports, industrial districts, large parks, protected natural areas, or rural landscapes where limited walkability reflects intended land uses rather than deficiencies in community design.

Overall, the National Walkability Index should be used alongside local knowledge, field observations, pedestrian infrastructure inventories, safety analyses, accessibility audits, and community engagement. While the index effectively measures characteristics of the built environment, it does not directly evaluate sidewalk quality, lighting, traffic speeds, crime, shade, topography, or other factors that influence the pedestrian experience.

Understanding Neighborhood Walkability

Neighborhood walkability reflects the degree to which residents can safely and conveniently reach everyday destinations on foot. Unlike municipal-level analyses, neighborhood-level measurements reveal localized differences in the built environment that are often hidden when entire cities are evaluated using a single average score.

The National Walkability Index measures four characteristics that research has consistently linked to walking behavior:

• Intersection Density, representing street network connectivity.
• Proximity to Transit Stops, measuring access to public transportation.
• Employment Mix, reflecting the diversity of employment opportunities.
• Employment and Household Mix, representing the balance between jobs and residential development.

Together, these variables estimate how supportive a neighborhood’s built environment is for walking. Numerous studies within the urban planning literature—including Cervero and Kockelman (1997), Ewing and Cervero (2010), and Frank et al. (2005)—have found that neighborhoods combining connected street networks, diverse land uses, and nearby destinations consistently generate more walking than neighborhoods characterized by disconnected streets and single-use development.

It is important to recognize that the National Walkability Index measures potential walkability rather than the quality of the pedestrian experience itself. Two neighborhoods with similar scores may differ substantially in sidewalk conditions, street trees, lighting, traffic safety, ADA accessibility, or perceived personal safety. Consequently, the index should be viewed as a planning indicator rather than a complete assessment of pedestrian conditions.

Perhaps the greatest strength of neighborhood-level analysis is its ability to identify variation within cities. Downtown districts, historic neighborhoods, university areas, and transit corridors frequently receive substantially higher scores than suburban neighborhoods located only a few miles away. This level of detail helps planners prioritize investments where improvements are likely to produce the greatest increases in accessibility and mobility.

Geographic Patterns in Neighborhood Walkability

Across the 2,740 neighborhoods analyzed, National Walkability Index (NWI) scores range from 1.00 to 20.00, with an overall mean of approximately 13.34 and a median of approximately 13.93. Using the seven fixed intervals applied in this analysis, the greatest concentrations occur between 11.43–14.29 (1,049 neighborhoods) and 14.29–17.14 (1,014 neighborhoods). Together, these two categories contain 2,063 neighborhoods, or approximately 75.3 percent of all neighborhoods, demonstrating that the study area is dominated by places with moderate-to-relatively-high support for walking rather than by either extreme.

This concentration is consistent with the structure of the EPA National Walkability Index. The NWI does not measure sidewalks or pedestrian conditions directly; instead, it evaluates characteristics of the built environment associated with the likelihood of walking. Specifically, the index incorporates street intersection density, proximity to transit, employment mix, and employment-and-household mix. Higher intersection density, shorter distances to transit, and greater diversity of land uses are associated with greater support for walking. The Smart Location Database similarly organizes the built environment around density, diversity, design, transit accessibility, and destination accessibility, reflecting the broader planning literature concerning the relationship between urban form and travel behavior.

The highest-scoring neighborhoods are disproportionately associated with downtowns, historic districts, compact mixed-use centers, and other highly connected urban environments. Examples include Charlottesville Historic District–Starr Hill (20.00), Denver Civic Center (19.67), Minneapolis Nicollet Island–East Bank (19.67), Eugene Downtown (19.58), Philadelphia Center City East (19.33), Philadelphia Old City (19.08), and San Francisco Chinatown (19.17). The pattern closely resembles EPA’s city-center example, where high intersection density, strong transit proximity, and mixed uses produce a NWI score of 17.5.

Downtowns are particularly revealing. Among the 52 neighborhoods containing “Downtown” in their names, the average NWI score is approximately 15.64, considerably higher than the dataset-wide average of 13.34. However, downtown status does not automatically produce high walkability. Downtown Waco scores only 8.71, while Downtown Gary scores 9.88, demonstrating that the functional characteristics captured by the EPA variables matter more than the neighborhood label itself. A downtown with limited transit accessibility, weaker land-use diversity, or lower street connectivity can therefore score substantially below a compact, transit-rich downtown.

A general center-to-edge pattern remains visible in many cities, but the data suggest that it should be described as a tendency rather than a universal rule. Compact central neighborhoods frequently outperform peripheral, low-density, or specialized areas because traditional urban centers tend to combine connected street networks, multiple destinations, employment, housing, and transit. EPA’s own examples progress from a low-scoring rural environment, through suburban residential development, to a historic main street and finally a city/suburban town center, illustrating how differences in built form correspond to increasing NWI values. EPA’s suburban example further demonstrates that a neighborhood can possess adequate pedestrian infrastructure yet receive a relatively low NWI score when transit accessibility and land-use mix are weak.

At the lower end of the distribution, specialized or sparsely developed landscapes appear prominently. Examples include New York City’s Jamaica Bay (West) (1.00), Los Angeles’ Chatsworth Reservoir (2.00), Jamaica Bay (East) (2.00), Boston’s Harbor Islands (2.67), Riverside’s Sycamore Canyon Park (3.00), Stockton’s Mariposa Lakes (3.00), and New Orleans’ Lake Catherine (3.67). These results require careful interpretation. Low NWI values in reservoirs, islands, parks, open spaces, airports, industrial districts, and similar environments do not necessarily represent planning failures. Instead, many such areas lack the mixture of residences, employment, connected streets, and transit accessibility that the NWI is specifically designed to reward. EPA similarly notes that rural areas can score poorly because destinations are dispersed, street networks are spread out, and transit access is absent.

The data also demonstrate substantial variation within individual municipalities, reinforcing the importance of neighborhood-scale analysis. New York City ranges from 1.00 to 17.22, Los Angeles from 2.00 to 17.40, Pittsburgh from 3.83 to 18.94, New Orleans from 3.67 to 18.67, and Boston from 2.67 to 17.61. These enormous internal ranges demonstrate that a single citywide walkability value can conceal fundamentally different built environments. This neighborhood-scale approach is particularly compatible with the EPA methodology because the original NWI is calculated for Census block groups rather than entire municipalities, providing a nationwide framework for comparing relatively small geographic units.

Overall Pattern

Taken as a whole, the results show that neighborhood walkability exists along a broad continuum, but the distribution in this dataset is strongly concentrated toward the center and upper-middle portions of the NWI scale. Of the 2,740 neighborhoods, 4 fall between 0–2.86, 30 between 2.86–5.71, and 209 between 5.71–8.57. By comparison, 295 fall between 8.57–11.43, 1,049 between 11.43–14.29, 1,014 between 14.29–17.14, and 139 between 17.14–20.00. Thus, extremely low and extremely high scores are comparatively uncommon, while intermediate and moderately high values dominate the dataset.

The concentration is particularly pronounced in the two largest categories. The 1,049 neighborhoods in the 11.43–14.29 interval represent approximately 38.3 percent of the dataset, while the 1,014 neighborhoods in the 14.29–17.14 interval account for approximately 37.0 percent. Combined, these 2,063 neighborhoods represent approximately 75.3 percent of all 2,740 neighborhoods analyzed. By contrast, only 243 neighborhoods, or approximately 8.9 percent, have NWI scores below 8.57, while 139 neighborhoods, or approximately 5.1 percent, occupy the highest interval of 17.14–20.00.

This distribution should not be interpreted as a direct measurement of how much people actually walk. EPA defines the NWI as a measure of relative walkability based on built-environment characteristics affecting the likelihood or feasibility of walking for utilitarian transportation. The component variables are ranked relative to other Census block groups, and the final index ranges from 1 to 20. Consequently, a neighborhood’s score indicates how strongly its built environment supports walking according to the EPA methodology rather than providing a direct measure of pedestrian activity, sidewalk quality, pedestrian comfort, or resident travel behavior.

The geographic results nevertheless show a clear relationship between urban structure and relative walkability. The strongest scores generally occur where several supportive characteristics coincide: connected street networks, diverse land uses, concentrations of housing and employment, and access to transit. Conversely, the lowest values tend to occur where one or more of these elements are absent, particularly in geographically isolated, undeveloped, automobile-oriented, or specialized-use environments. This finding closely follows the logic of the EPA methodology, which combines land-use diversity, street connectivity, and transit access into a composite measure of the built environment’s support for walking.

The results also demonstrate that walkability is more accurately understood as a neighborhood-scale characteristic than as a uniform municipal characteristic. A city can simultaneously contain some of the most and least walkable environments in the dataset. The extreme internal variation found in New York City, Los Angeles, Boston, Pittsburgh, and New Orleans illustrates how municipal boundaries encompass downtowns, traditional residential neighborhoods, suburban districts, industrial areas, parks, airports, waterways, and other land uses with fundamentally different spatial characteristics. Citywide averages therefore provide useful summaries but can obscure the geography that actually produces walkability.

Finally, the seven-class analysis adds an important distinction to the EPA’s original four-category classification. EPA officially divides NWI scores into Least Walkable (1–5.75), Below Average Walkable (5.76–10.5), Above Average Walkable (10.51–15.25), and Most Walkable (15.26–20). The seven equal intervals used in this analysis provide greater analytical resolution within that 1–20 continuum, making differences between neighborhoods more visible. They should therefore be presented as analytical subclasses derived from the EPA index, rather than as replacements for EPA’s official categories.

Overall, the results reinforce the central premise of the National Walkability Index: walkability emerges from the interaction of multiple characteristics of the built environment rather than from any single feature. Compact urban centers frequently perform well because connectivity, land-use diversity, employment, housing, and transit reinforce one another, while dispersed or specialized landscapes generally perform less strongly. At the same time, the substantial variation both between and within cities demonstrates why neighborhood-scale GIS analysis is particularly valuable. It reveals spatial differences that municipal averages cannot capture and provides planners with a more precise basis for identifying where changes to connectivity, land-use patterns, and transit accessibility may strengthen the built environment’s support for walking.