Environmental audits are used to assess supports for physical activity in the community. Understanding the suitability of such instruments for use by community members is crucial for advocacy and participatory research. This study examined the reliability of an audit instrument filled out by trained researchers and untrained community members. Two researchers and five community members conducted environmental audits on a total of 335 street segments in lower-income areas in St Louis, Missouri (representing a "low-walkable city"), and Savannah, Georgia (representing a "high-walkable" city). The audit tool consisted of six major sections--land use environment, recreational facilities, transportation environment, aesthetics, signage, and social environment. Interrater agreement between researchers and community members was assessed using percent observed agreement and the kappa statistic. According to observed agreement, the majority of audit items (67 of 76) had substantial to almost perfect agreement (> or =0.60) between researchers and community members. However, much lower agreement was observed using the kappa statistic (only 8 of 76 items with kappas > or =0.60). With some formal training, this audit tool may be useful for advocacy and participatory research to assess the activity friendliness of neighborhood environments.
