[en] Purpose: We have previously shown that the distribution of PSA doubling times (PSADT) varies widely in newly diagnosed prostate cancers. This report examines PSADT as an independent variable predictive of biochemically free of disease (bNED) survival. Materials and Methods: One hundred and eight patients with T1-3 NX prostate cancer treated between (2(89)) and(11(93)) have PSADT's calculated from 3 or more PSA values obtained over 60 days or more (median 9 months) prior to treatment. Multivariate analysis of bNED is performed with two groups of PSADT (<12 months and 12+ months) and with three groups (<12 months, >12 months, not rising) in addition to common prognostic indicators (dose, technique, and Gleason Score). bNED failure is PSA ≥1.5 ngm/ml and rising on two consecutive values. Results: Figure 1 illustrates the significant differences in bNED survival between the 3 PSADT groups (p=.007). Table 1 presents the results of multivariate analyses for the two PSADT groups and the three PSADT groups. PSADT is the strongest independent variable in both analyses (p=.02 and p=.01, respectively) and center of prostate dose is also significant (p=.05 and p=.04, respectively). Gleason Score and technique are not significant in either analysis. Conclusion: 1. PSADT is shown to be an independent predictor of bNED survival in prostate cancer treated with external beam radiation. Pretreatment PSADT and post radiation failure PSADT are now both independent predictors of subsequent outcome. 2. Patients with PSADT of <12 months have aggressive disease and should be considered for aggressive treatment programs. 3. A prolonged or not rising PSADT may be one factor considered in evaluating prostate cancer patients for surveillance. 4. Pretreatment PSADT may be particularly useful in patients where the pretreatment PSA level does not reflect the behavior of the cancer (low PSA with short PSADT; high PSA with long PSADT)