This paper reviews the reliability and validity of an instrument for assessing health-related quality of life (HRQOL) in CaPSURE (Cancer of the Prostate Strategic Urologic Research Endeavor), an observational database of men with biopsy-confirmed prostate cancer. The questionnaire includes the RAND 36 item Health Survey 1.0, the UCLA Prostate Cancer Index and items measuring self-esteem and the impact of prostate cancer in general and on the family. The reliability and validity of this instrument are reported for a group of 2,382 men enrolled in CaPSURE. Individuals complete HRQOL questionnaires at enrollment and quarterly thereafter. The subscales of the instrument resulted in a high internal-consistency reliability across all subscales (range = 0.75-0.94). The test-retest reliability was good with the exception of the comparative health item (intraclass correlation coefficient (ICC) = 0.55). The concurrent validity data included moderately strong associations with subscales of similar concepts and the ability to distinguish between patients who are known to be different--newly diagnosed (within 6 months) and patients diagnosed and treated an average of 3 years ago. Overall, this instrument demonstrated good reliability and validity and supported the need for incorporating HRQOL as a component of outcomes assessment in the management of patients with prostate cancer.