The potential for physical activity (PA) to prevent and treat chronic illnesses is well known. Because of increases in the numbers of elderly and because the elderly are the heaviest users of health care services, efforts to prevent and treat chronic illnesses are paramount. PA is a low cost treatment modality and has been demonstrated to benefit seniors by facilitating the maintenance independent living arrangements, reducing the risk of falling and fracturing bones. Additionally PA has also been demonstrated to be effective in reducing blood pressure, improving balance and lessening age-related insulin resistance that can lead to Type II diabetes mellitus. Unfortunately data indicate that seniors are among the most sedentary Americans. Thus efforts to create more physically active seniors are advised. Knowledge of environmental and demographic determinants of PA could be valuable information in designing PA interventions for this group. This study sought to identify factors that may influence physical activity. Retirement village residents, living in Mesa, Arizona were randomly recruited to complete a 55-item questionnaire and the PASE physical activity recall instrument. The questionnaire requested information on individual's abilities to perform household and leisure activities, perception of current physical condition, smoking behavior, demographic information and the detailed PASE. The PASE produces a summary index that totals light, moderate and strenuous PA. The PA environment was assessed by 4 items that measure the ease of crossing streets, pleasantness of the neighborhood, driver behavior and overall safety level of the neighborhood. Linear regression was utilized to assess predictors of PA. Of 250 questionnaires distributed, 113 (45.2%) were returned to the investigator. Mean respondent age was 70.9 (7.71) with a range of 51 to 91 years, mean BMI was 26.3 (3.50) with a range of 19.2 to 37.3. Females outnumbered males 64 to 49. Only 2.6% of respondents smoked and all participant ethnicities were self-reported as White. The mean and standard deviation PASE scores for males was 37.86 (25.6) and for females 25.4 (25.0). This difference was statistically significant F (1(4.38) p=.04). The regression model was significant (p=.005) and explained 16.5% of the variance in PA. Significant predictors of PA were BMI, age and gender. As BMI and age increased PA decreased and males reported more PA than females. Perception of the neighborhood failed to predict PA. All participants lived in the same retirement village and scores describing that physical environment reasonably uniform, x=4.9(0.74). The present model only identified demographic variables as significant predictors of PA.