Seventy-five cirrhotic patients with hyperammonemia in the past or at the time of the study were randomly divided into two groups (treated with lactulose or nontreated) in 14 hospitals in Japan. Thirty-six cirrhotic patients were diagnosed as having subclinical hepatic encephalopathy (SHE), and 39 were diagnosed as non-SHE. SHE was diagnosed when the results of all three of the quantitative psychometric tests used (number connection test, and symbol digit and block design tests of the Wechsler adult intelligence scale [revised]) were abnormal as compared with age-matched normal values. The mean number of abnormal psychometric test results and the prevalence of SHE were used for a quantitative evaluation of the efficacy of the lactulose treatment. Twenty-two of the SHE patients were treated with lactulose (45 mL/d) for 8 weeks, and the other 14 SHE patients did not receive lactulose. In the SHE patients administered lactulose, the results of the quantitative psychometric evaluation were significantly improved at 4 and 8 weeks after the beginning of the lactulose administration. The SHE had disappeared in 10 (50%) of the 20 treated patients at week 8, but it persisted in 11 (85%) of the untreated 13 patients. We concluded that lactulose treatment in cirrhotic patients with SHE is effective with respect to psychometric tests.