Background: Web-based worksite interventions represent a feasible, acceptable, and cost-effective approach to address the physical inactivity burden among nurses. We delivered a 6-week web-based worksite intervention to promote physical activity (PA) among nurses, and assessed whether demographic (age, sex), occupational (weekly hours, shift schedule, work role), and psychological (mood states) factors predicted changes in weekly moderate-to-vigorous intensity PA (MVPA) and steps. Methods: Nurses (N=76; 97% female; Mage = 46±11 years) from the University of Ottawa Heart Institute were recruited into this parallel-group trial and randomized to an individual, friend, or team PA challenge group. Pre-randomization, nurses completed self-report questionnaires and were given a PA monitor to wear prior to and during the intervention. We analyzed data using multilevel modeling for repeated measures. Results: Changes in weekly MVPA were significant and followed an inverted U-shape. Changes in weekly steps were also non-linear, but non-significant. Shift schedule (rotating vs. fixed) by time (estimate=-17.62, SE=5.39) and by time-squared (estimate=2.66, SE=0.86) and work role (clinical only vs. other) by time (estimate=19.47, SE=5.69) and by time-squared (estimate=-3.11, SE=0.91) predicted MVPA. Work role by time (estimate=1096.85, SE=522.50) and by time-squared (estimate=-218.40, SE=97.31) predicted step count. Age, sex, weekly hours, and mood states did not predict weekly MVPA or steps. Conclusions: Strategies should be employed to help nurses maintain increases in PA observed during a web-based worksite intervention. As nurses working rotating shifts and completing clinical work only showed less improvements in PA, collaborating with such nurses to inform the design of future interventions is recommended.