Patients with kidney stones are routinely advised to increase their fluid intake to decrease the risk of stone recurrence. However, there has been no detailed examination to determine whether the effect on recurrence varies by the type of beverage consumed. The authors conducted a prospective study of the relation between the intake of 21 different beverages and the risk of symptomatic kidney stones in a cohort of 45,289 men, 40-75 years of age, who had no history of kidney stones. Beverage use and other dietary information was measured by means of a semiquantitative food frequency questionnaire in 1986. During 6 years of follow-up (242,100 person-years), 753 incident cases of kidney stones were documented. After adjusting simultaneously for age, dietary intake of calcium, animal protein and potassium, thiazide use, geographic region, profession, and total fluid intake, consumption of specific beverages significantly added to the prediction of kidney stone risk (p < 0.001). After mutually adjusting for the intake of other beverages, the risk of stone formation decreased by the following amount for each 240-ml (8-oz) serving consumed daily: caffeinated coffee, 10% (95% confidence interval 4-15%); decaffeinated coffee, 10% (3-16%); tea, 14% (5-22%); beer, 21% (12-30%); and wine, 39% (10-58%). For each 240-ml serving consumed daily, the risk of stone formation increased by 35% (4-75%) for apple juice and 37% (1-85%) for grapefruit juice. The authors conclude that beverage type may have an effect on stone formation that involves more than additional fluid intake alone.