Recent studies report that certain dietary patterns, especially those high in red and processed meats, are associated with prostate cancer risk. We prospectively investigated associations between empirically derived dietary patterns and prostate cancer risk using the Melbourne Collaborative Cohort Study. We followed 14,627 men of ages 34 to 75 years for an average of 13.6 years, and identified 1,018 incident prostate cancers. Factor analysis of the 121-item food frequency questionnaire identified four factors with eigenvalues >2 that explained 67% of the total variance. Using Cox proportional hazard models, we found no association between any dietary pattern and prostate cancer risk overall (all P(trend) >or= 0.2). The hazard ratios for quartiles of the dietary pattern scores ranged from 0.87 to 1.14 and all 95% confidence intervals included 1. The analyses by aggressiveness, Gleason score groups, and age at diagnosis did not show any association (all P(trend) > 0.07).