The majority of U.S. adults perform insufficient amounts of physical activity to prevent disease and maintain fitness. National recommendations prescribe fixed physical activity goals (e.g. 10,000 steps per day) that may fall outside of an individual's current physical activity repertoire. Prescribing smaller, more adaptive goals based on participant past behavior may be more efficacious at increasing physical activity to the target level. This study tested a pedometer-based intervention that prescribed adaptive goals and rewarded behavior using a percentile schedule of reinforcement. Five individuals enrolled into the intervention and were evaluated with a single-case withdrawal (ABA) design over 10 weeks. The six-week intervention consisted of one-time educational materials, daily adaptive goals, and contingent financial rewards administered on a percentile schedule. Daily goals were determined by ranking a participant's prior 9 days of physical activity (i.e. step counts) and selecting the 40 ^th percentile of the distribution on a moving basis. A Lifecorder Plus, combined accelerometer and pedometer, measured moderate-to-vigorous physical activity (MVPA) minutes per day and steps per day simultaneously. Visual analyses and multilevel statistical models for longitudinal data tested for change across phases. Based on visual analysis, four of the five women increased their median number of steps/day, and all five increased their median MVPA minutes/day. Participants increased their activity by 851 steps/day (range -829 to 2,450 steps) or approximately 5,957 steps per week, and 3.34 MVPA minutes/day (range 1.93 to 17.27 minutes) or approximately 23.38 MVPA minutes per week from baseline to the intervention phase. After controlling for covariates, the multilevel model detected a significant increase of 551.21 steps/day (SE = 258.26, p = .03) and 2.65 MVPA minutes/day (SE = 1.09, p = .02) during the intervention phase compared to the baseline phase, after adjusting for wear time and day of the week. This study provides a formal test of percentile schedules for physical activity research and provided intervention efficacy (i.e., ‘proof of concept’). The findings may be used as a preliminary study to inform future work in this line of research.