A growing body of evidence supports a positive relationship between descriptive norms and physical activity using both correlational (Carpenter & Amaravadi, 2019) and experimental designs (Crozier, 2019). Although the findings align with conceptual models (Cialdini et al., 1990), this relationship becomes problematic if the desired behavior is not the norm. Consider physical activity, where only 18% of Canadian adults are active enough for health benefits. What can be done if the intended behavior (i.e., being physically active) counters the prevailing norm (i.e., physical inactivity)? One possibility concerns the use of dynamic norms (Sparkman & Walton, 2017), which provide information that a behavior is changing. This study examined whether dynamic norms would increase interest in exercising more during final examinations in students who were regular exercisers but indicated they reduced exercise during the exams. Students were assigned randomly to one of four conditions: Dynamic norm (n=28), dynamic norm/high outcome expectation (n=34), minority descriptive norm (n=34), or attention control (n=34). Interest in increasing exercise during the exams was the main dependent variable. ANCOVA results revealed a significant condition effect (p=0.007, ?p2=0.09). Post hoc tests revealed interest in increasing exercise during exams was higher in both dynamic norm conditions (ps <.03) versus attention control. No difference emerged between the minority descriptive norm and attention control (p=0.84). The current findings provide initial evidence that exposure to dynamic norms may provide students with the impetus to increase exercise during an examination period even though the prevailing norm may serve as a barrier to change.