The study purpose was to determine whether unsealed (with participant recording of daily steps) versus sealed pedometer monitoring elicits reactivity (i.e., changed behavior due to awareness of being monitored) in adults. In a randomized controlled crossover study, 28 healthy White adults (12 men, M age = 51 ± 9.9 years; and 16 women, M age = 44 ± 6.5 years) completed 2 assessment types (sealed versus unsealed pedometer). Half of the participants wore a sealed pedometer for 1 week, submitting it for reading by research staff. After a 2-week washout, the participants wore an unsealed pedometer, recording total daily steps for 1 week. The remaining participants completed the protocol in reverse order. Analyses included a repeated measures analysis of variance (ANOVA) using condition (mean steps per day for sealed versus unsealed) as the repeated variable and gender and order of assessment type as the between variables. There was no significant difference between sealed versus unsealed pedometers in steps per day, F( 1,24) = 0.51, p = .48, regardless of gender, F(1, 24) = 0.13,p = .72, or order of assessment type, F(1,24) = .48, p = .50. The intraclass correlation coefficient (ICC) between average steps per day for the sealed versus unsealed assessment was .84, p < .0001, confirming strong repeatability (i.e., individuals hold their rank order within the group regardless of condition type). Reactivity does not appear to be an important threat to validity of pedometer self-monitoring in free-living healthy adults. Copyright © 2006, Lawrence Erlbaum Associates, Inc.