Purpose: To evaluate the safety and activity of 6 months of treatment with lenalidomide at 5 or 25 mg/d in nonmetastatic biochemically relapsed prostate cancer. Experimental Design: Sixty men with non-castrate, nonmetastatic, biochemically relapsed prostate cancer were stratified by prostate-specific antigen (PSA) doubling time, surgery/radiation therapy, prior androgen deprivation therapy (ADT), and randomized to lenalidomide 5 mg (n = 26) or 25 mg/d (n = 34) for 3 weeks repeated monthly for 6 months or until dose-limiting toxicity or disease progression. Toxicity was evaluated monthly, and PSAs and X-rays/scans every 6 months. Study size was determined to detect a progression rate of 40% at 6 months in either arm with 85% power (compared with a rate of 80% in the population receiving no treatment). Changes in PSA slopes were calculated using the regression of the log PSA for each patient before and during the initial 6 months and compared by t test. Results: Baseline variables were balanced between arms. Grade 3/4 toxicity rates were 12% (n = 3) with 5 mg and 29% (n = 10) with 25 mg (P = 0.1), most commonly neutropenia (five patients, all on 25 mg). Two patients per arm had thromboembolic events. The change in PSA slope was greater with 25 mg versus 5 mg [-0.172 (-0.24 to -0.11) versus -0.033 (-0.11 to 0.04); P = 0.005]. With a mean follow-up of 31.4 months (range 14-44), five patients on 25 mg and one patient on 5 mg remain on the study. Conclusions: Lenalidomide has acceptable toxicity and is associated with long-term disease stabilization and PSA declines. Randomized studies evaluating conventional clinical disease end points in this patient population are planned.