Technical Abstract: PURPOSE: To evaluate the validity of the Previous Day Physical Activity Recall (PDPAR) as a physical activity diary in adolescents using two accelerometer intensity classifications. METHODS: One hundred eighth graders (47 boys, 53 girls) used the PDPAR as a daily diary and wore MTI accelerometers for four consecutive days. Measured time spent in moderate (> or = 3 METs) and vigorous (> or = 6 METs) activity was based on two published MTI cut-point limits (that of Freedson et al./Trost et al. and that of Puyau et al.). Spearman rank order correlations and Bland-Altman plots were used to examine agreement between MTI and PDPAR diary estimates of activity. RESULTS: MTI estimates of mean minutes per day of total moderate to vigorous physical activity (MVPA) were 65.2 (+/-43.2) using the Freedson et al./Trost et al. cutoffs and 17.5 (+/-18.5) using those of Puyau et al., while students self-reported 105.1 (+/-80.1) min.d(-1). Significant relationships were observed between the diary and MTI for total MVPA using either the Freedson et al./Trost et al. (r = 0.42) or Puyau et al. (r = 0.41) cutoff as well as raw counts (r = 0.44). Plots showed reasonable agreement between the diary and Freedson et al./Trost et al. MTI estimates of MVPA for daily totals of < or = 60 min, but the Puyau et al. estimates were consistently lower. Diaries overestimated activity as time increased when compared to either MTI cut point, especially on vigorous activity. CONCLUSIONS: Time estimates of MVPA differed by assessment tool, but diary estimates showed adequate association with the MTI. Diaries reflected intensity-specific activity, corresponding most closely with the Freedson et al./Trost et al. classification of moderate, but substantially overestimated vigorous activity regardless of cut-point method. This is likely due to the measurement characteristics of the PDPAR, which classifies activities in 30-min blocks, as well as the nature of common activities in which high levels of intensity are not sustained.