To assess whether patients can weigh risk comparisons involving mortality and quality of life in an understandable manner based on their willingness to accept risks of complications.Cross-sectional survey of patients.University-based Department of Veterans Affairs Medical Center.230 men patients seen in a general medicine clinic.Two survival graphs were used. Each graph contained survival curves for two alternative unidentified treatments for an unidentified medical condition. Graph 2 contained one curve that had a life expectancy that was 14% higher than the life expectancy of the corresponding curve in graph 1. Respondents were randomly assigned one of the two graphs and were asked to indicate which treatment they preferred and what risk of a change in their quality of life (urinary incontinence or importance) they were willing to accept to achieve longer survival. Patients were also asked whether they had a history of urinary incontinence or impotence.Patients tended to be unwilling to accept worse quality of life to achieve increased survival over time. For both curve comparisons, significantly more (p < 0.01) patients accepted a treatment associated with higher mortality to avoid a 100% chance of incontinence than to avoid a 100% chance of impotence. Of the 75% (172/230) of patients reporting willingness to accept risk of either urinary incontinence or total impotence or both, 62% reported having at least some symptoms related to urinary incontinence or impotence. Of the 58 patients not willing to accept the complication risks, only 11% reported a history of urinary incontinence or impotence. The results show that patients are able to make distinctions about severity of morbidity, men are less willing to accept the risk of urinary incontinence than that of total impotence, and men patients who are symptomatic with urinary incontinence or impotence are more willing to accept the risks of treatment than are asymptomatic patients.