All forms of androgen-deprivation therapy, including finasteride, induce distinctive histologic changes in benign and neoplastic prostatic epithelial cells, including cytoplasmic clearing, nuclear and nucleolar shrinkage, and chromatin condensation. Treated cancer has a significantly higher architectural (Gleason) grade, lower nuclear grade, and smaller nucleolar diameter than untreated controls, creating the potential for grading bias. Recognition of these changes may be difficult in needle biopsies and lymph node metastases with treated cancer because of the subtle infiltrative pattern and inconspicuous nucleoli. The effects of finasteride may be less pronounced than other forms of therapy with variable distribution throughout the prostate; further, there may be greater sensitivity of low and intermediate-grade cancer than high-grade cancer. The Gleason grading system for cancer should not be used after finasteride treatment as it is not validated in this setting and is likely to overestimate the biologic potential of high-grade cancer observed after therapy. Chemoprevention trials with agents such as finasteride that alter morphology should not rely on cancer grading as a secondary endpoint owing to grading bias.