It has been suggested that aspects of lactose consumption and metabolism favoring a relatively high tissue level of galactose-1-phosphate may predispose women to ovarian cancer. The authors sought to examine this hypothesis in a study of 108 18- to 74-year-old Caucasian residents of a three-county area of western Washington who were diagnosed with stage I ovarian cancer during 1989-1991, and 108 age- and race-matched controls. Lactose and galactose intake, measured using a food frequency questionnaire, had been hypothesized to increase risk, but were somewhat lower among the cases than among the controls (75th percentile of lactose intake vs. 25th: odds ratio (OR) = 0.80, 95% confidence interval (Cl) 0.52-1.2; of galactose intake: OR = 0.71, 95% Cl 0.48-1.1). Intestinal lactase activity, also hypothesized to have a positive relation with ovarian cancer occurrence, was measured with an oral lactose challenge followed by determination of urinary galactose; no evidence that it was related to the disease was found (75th percentile of excreted galactose vs. 25th: OR = 0.87, 95% Cl 0.62-1.2). Galactose-1-phosphate uridyltransferase (transferase), the enzyme responsible for the metabolism of galactose-1-phosphate, was measured in erythrocytes; no deficit in cases was observed (75th percentile of transferase activity vs. 25th: OR = 1.3, 95% Cl 0.80-2.1). There was also no excess of cases carrying low-activity genetic variants of the transferase enzyme (lower-activity variants vs. higher-activity variants: OR = 0.61, 95% Cl 0.21-1.7). These results do not support the hypothesis that aspects of lactose and galactose intake and metabolism have a bearing on the etiology of ovarian cancer.