This study identified factors that predicted improved levels of physical activity in working women. Data were from a prospective, controlled trial. Sedentary women (N = 287) were recruited during an employer-sponsored health screening. The experimental group received a brief, tailored physical activity intervention as well as a booster phone call 2 weeks later. Participants were evaluated by comparing pre-test/post-test measures at 6-week follow-up. It was hypothesized that variables in the Transtheoretical Model (baseline stage of change, perceived pros, cons, and self-efficacy), as well as demographic variables and physical characteristics would predict improved physical activity. In addition, the study tested whether increased physical activity changed the variables of the Transtheoretical Model. Factors that correlated to improved physical activity were entered into regression models to determine predictors of improved physical activity. Being assigned to the intervention group was a significant predictor in all of the regression models. Women who increased their stage of change of physical activity had greater improvement in self-efficacy (p = .013). Minority women showed greater improvement in stage of change of physical activity (p = .001) as well as week-end physical activity (p = .042). Women with less education were more likely to increase total minutes walked per week (p = .038). The research suggests that these cognitive and demographic variables are important for improving levels of physical activity. The Transtheoretical Model was partially supported by the analyses.