The role of physical activity (PA) in preventing chronic disease is well documented. The purpose of this study was to determine the relationship of PA levels with personal characteristics among a college student cohort. Undergraduates (n=1700) were randomly selected and mailed an 89-item questionnaire adapted from the CDC's National College Health Risk Behavior Survey (NCHRBS, 1997). Students were queried about PA and a variety of health behaviors. Chi-square analysis was done to determine associations among variables; logistic regression was conducted to determine predictors of PA. Statistically significant results are presented here. Of the 903 respondents, 61% were females, 90% were white-non-Hispanic. Mean age was 20 (+ 1.55). Mean BMI for males was 24.0 (+ 4.1), for females was 23.0 (+ 4.2). Nearly 46% of respondents indicated they "Exercised or participated in sports activities for at least 20 minutes that made you sweat or breathe hard" on three or more of the preceding 7 days (VIG) while 59% reported they "Walked or biked for at least 30 minutes at a time" on 3 or more of the preceding 7 days (MOD). VIG was significantly associated with being in a sorority or fraternity, trying to change current weight, perceiving being overweight or obese (p<.001 in each case), and being a male who has been drunk in the last 30 days (p<.05). MOD was associated with trying to change current weight (p<.001) and being female and living on campus (p<.05 in both cases). Logistic regression analysis indicated BMI, being in a sorority or fraternity, being male, trying to change weight and perceiving being overweight or obese all predicted VIG. Trying to change weight and living on campus predicted MOD. High levels of PA among the sample may be due to overestimation of self-reported PA. These results indicated males were more likely to be VIG than females, while females were more likely to be MOD. Trying to change weight was associated with both levels of activity while perceiving being overweight or obese was only associated with VIG. It is thought that engaging in one healthy behavior predicts other health behaviors yet VIG was associated with having been drunk in the previous 30 days. The association between MOD with living on campus is not surprising as this requires walking or biking for transportation. These results may help identify those more or less likely to engage in activity therefore more closely targeting interventions.