Abstract Background Choice architecture interventions, which subtly change the environment in which individuals make decisions, can be used to promote behavior change. This systematic review aimed to summarize studies on micro-environmental choice architecture interventions that encouraged physical activity or discouraged sedentary behavior in adults, and to describe the effectiveness of those interventions on these behaviors – and on related intentions or health outcomes – in presence of the intervention and after removal of the intervention (i.e. post-intervention, regardless of the time elapsed). Methods We systematically searched PubMed, Embase, PsycINFO and the Cochrane Library for (quasi) experimental studies published up to December 2019 that evaluated the effect of choice architecture interventions on physical activity and sedentary behavior, as well as on intentions and health outcomes related to physical activity/sedentary behavior. Studies that combined choice architecture techniques with other behavior change techniques were excluded. All studies were screened for eligibility, relevant data was extracted and two independent reviewers assessed the methodological quality using the QualSyst tool. Results Of the 9609 records initially identified, 88 studies met our eligibility criteria. Most studies (n = 70) were of high methodologic quality. Eighty-six studies targeted physical activity, predominantly stair use, whereas two studies targeted sedentary behavior, and one targeted both behaviors. Intervention techniques identified were prompting (n = 53), message framing (n = 24), social comparison (n = 12), feedback (n = 8), default change (n = 1) and anchoring (n = 1). In presence of the intervention, 68% of the studies reported an effect of choice architecture on behavior, whereas after removal of the intervention only 47% of the studies reported a significant effect. For all choice architecture techniques identified, except for message framing, the majority of studies reported a significant effect on behavioral intentions or behavior in presence of the intervention. Conclusions The results suggest that prompting can effectively encourage stair use in adults, especially in presence of a prompt. The effectiveness of the choice architecture techniques social influence, feedback, default change and anchoring cannot be assessed based on this review. More (controlled) studies are needed to assess the (sustained) effectiveness of choice architecture interventions on sedentary behavior and other types of physical activity than stair use.