BACKGROUND: Physical activity interventions using habit development may help people increase and then maintain physical activity increases over time. Enacting behavior in consistent contexts is a central component of habit development, yet its causal role in habit development in health behaviors has not been confirmed. PURPOSE: This study tests the causal role of consistent context in habit development in health behavior, using a randomized control trial of a planning intervention to develop a walking habit in 127 insufficiently active, working, midlife adults in a real-world setting. METHODS: We compare participants who plan walking in consistent contexts with controls who plan walking in varied contexts and with controls not required to plan on a change in average daily steps (measured using an accelerometer) and inhabit automaticity during a 4-week intervention and at a 4-week follow-up. RESULTS: As expected, consistent and varied context planners increased walking during the intervention, but only consistent context planners developed (and maintained) habit automaticity. Counter to expectations, consistent context planners did not show walking maintenance. However, across conditions, participants who developed more habit automaticity during the intervention also maintained walking more (decreased less). Having a routine daily schedule moderated some effects. Notably, no-plan controls with greater routine developed more habit automaticity, mediated by walking in more consistent contexts. CONCLUSIONS: This study confirms the causal role of consistent contexts in developing a walking habit, in a real-world setting, with an important but challenging population for physical activity interventions and identifies a facilitating condition common for many: a routine schedule. Developing an exercise habit may help people increase and then maintain physical activity. This study tests and confirms the role of exercising in consistent contexts as a cause of forming a daily walking habit. We use a randomized control trial of a 4-week planning intervention, with a follow-up 4 weeks after the intervention. Participants were 127 insufficiently active, working, midlife adults. We compared participants asked to plan their daily walks in consistent contexts from day-to-day, with participants asked to plan their walks in varied contexts and with participants not required to plan. As expected, consistent and varied context planners increased their daily walking steps (measured using an accelerometer) during the intervention compared to participants not required to plan. However, only consistent context planners developed (and then maintained) a daily walking habit, that is, where taking daily walks felt relatively automatic. Unexpectedly, consistent context planners did not show walking maintenance. However, across all participants, those who developed a stronger walking habit during the intervention maintained their walking more after the intervention ended. Lastly, having an existing routine daily schedule helped some participants. Those who were not asked to plan and had a more routine daily schedule also developed a daily walking habit.