Objective: To examine the uptake of cycling for recreation and transport, and relate these behaviors to individual, social, and environmental exposures over time. Method: Data were drawn from 909 adults in Time 2 (T2) (2005-2006) and Time 3 (T3) (2007-2008) of the RESIDE study (Australia). Demographics, perceptions of self-efficacy and social support related to cycling, neighborhood environment perceptions, and objective measures of the neighborhood were measured at T2. These were compared with uptake of cycling for recreation and transport at T3. Results: At T3, 54 (5.9%) had taken up cycling for recreation and 44 (4.8%) for transport. Positive perceptions of self-efficacy at T2 were consistently positively associated with the uptake of cycling for either purpose at T3. Respondents living in higher walkable neighborhoods (OR=1.63; 95% CI=1.02-2.62) or with higher objectively measured street connectivity (OR=1.80; 95% CI=1.05-3.07) were more likely to start cycling for recreation when compared with their reference groups. No significant relationships existed between objective measures of the neighborhood and uptake of cycling for transport. Conclusion: Interventions focusing on enhancing self-efficacy and generating social support will likely positively influence both cycling for recreation and transport; and providing infrastructure that creates physically supportive neighborhoods may increase cycling levels.