Objective: This brief report describes the possibility of reactivity in the baseline assessment of physical activity using accelerometry from two separate randomized controlled trials of a behavior intervention for increasing physical activity in persons with multiple sclerosis (MS). Method: The samples included 18 persons with multiple sclerosis (MS) from Study 1 and 20 from Study 2 who were randomized into treatment arms of the intervention. The participants initially wore an accelerometer over a 7-day period for collection of baseline data, and one week later wore a pedometer over a 7-day period for collection of data for self-monitoring and goal setting in week 1 of the 12-week intervention. The accelerometer and pedometer data were both expressed in average steps per day over a 7-day period. Results: There was a moderate (d = .56), statistically significant (p = .03) difference of 1,822 steps per day between baseline and week 1 of the intervention in Study 1. There was a large (d = 1.36), statistically significant (p = .0001) difference of 2,338 steps per day in Study 2. Conclusion: We are unaware of other research describing a significant change in physical activity between baseline and the first week of a behavioral intervention and believe that this change reflects reactivity in the baseline assessment using accelerometry.