Background: Youth physical activity is partly influenced by access to playgrounds and recreational opportunities. Playgrounds in disadvantaged areas may be less safe. Methods: Investigators assessed safety at 154 playgrounds in Boston between July 2000 and July 2001. Playgrounds were geocoded and safety scores assigned to census block groups (CBGs). For each of Boston's 591 CBGs, investigators calculated the total number youth and proportions of black residents, adults without a high school degree, and youth living in poverty. Investigators assigned each CBG a safety score, and calculated distance from the CBG centroid to the nearest playground and nearest "safe" playground (top safety quartile). Statistical analyses were completed using SAS PROC GENMOD by October 2002. Results: In bivariate analysis, playground safety was inversely associated with total CBG youth population (p =0.001) and proportions of black residents (p <0.001), youth in poverty (p =0.003) and residents with no high school degree (p <0.002). The proportion of black residents in the CBG was inversely associated with safety (p =0.013), independent of CBG educational attainment and numbers of youth. The average distance was 417 meters to the nearest playground and 1133 meters to the nearest "safe" playground. Distance to the nearest playground was inversely associated with the proportion of residents with no high school degree (p <0.0001) after controlling for numbers of youth and proportion of black residents. CBGs with more youth had greater distances to the safest playgrounds (p =0.04). Conclusions: In Boston, playground safety and access to playgrounds varied according to indicators of small-area socioeconomic and racial/ethnic composition.