The PCPT is a chemoprevention trial of finasteride with a primary endp
oint of biopsy-proven presence or absence of prostate cancer. A total
of 18,000 healthy men, aged 55 years and older, will be randomized. Ha
lf will receive finasteride (5 mg/day) and half will receive placebo (
one matching tablet per day) for 7 years. The trial is designed to hav
e 92% power to detect a 25% reduction in period prevalence of biopsy-p
roven disease using a two-sided test with alpha = 0.05. The trial is c
omplicated by the known impact of finasteride on the major screening t
est for prostate cancer, prostate specific antigen (PSA). This paper d
escribes the PCPT design with reference to alternatives that were cons
idered. The chosen design depends on five critical assumptions that mu
st be monitored closely throughout the 9-year trial.