PURPOSE: Physical activity (PA) diaries reduce the recall error inherent in self-reported PA but are burdensome. The purpose of this study was to compare a cell phone-based diary with a paper diary and examine the reliability and validity of the cell phone diary. METHODS: In a pilot study, 25 women and 23 men, age 45-65 yr, completed cell phone and paper PA diaries 4 d.wk(-1) for three consecutive weeks and a user satisfaction survey. In the subsequent validation study, 623 middle-age participants (52.5% women) were asked to complete the cell phone diary and wear an accelerometer for two 7-d periods, approximately 6 months apart. They also completed two PA questionnaires. Fitness, body mass index, and percent body fat were obtained as indirect validation criteria. RESULTS: Estimates of PA from the cell phone and paper diaries were similar (mean within person difference = -43.8 MET.min.d(-1) of total PA, SD = 360, P = 0.49, 7.4 min.d(-1) of moderate-vigorous PA, SD = 66, P = 0.53). Users preferred the cell phone diary over the paper diary (59.6% vs 35.4%). In the subsequent study, intraclass correlations for the cell phone diary ranged from 0.55 for light PA to 0.63 for vigorous PA. Although PA estimates from the cell phone diary were generally significantly higher than those from the accelerometer and the questionnaires, correlations for moderate and vigorous PA were moderate (rho = 0.25-0.59 with the questionnaires and 0.27-0.35 with the accelerometer). The correlations between the cell phone diary and the indirect validation criteria were generally in the expected direction and of moderate magnitude. CONCLUSIONS: A cell phone-based PA diary is equivalent to a paper diary, acceptable to users, and a relatively reliable and valid approach to self-reported PA.