The prevalence and the clinical implications of subclinical cognitive alterations in cirrhotic patients have not been well defined as yet. Therefore, we performed a study to assess the clinical features and the survival of cirrhotic patients with cognitive alterations detected by the number connection test (NCT) and a set of computerized psychometric tests (Scan, Choice1, and Choice2) measuring the reaction times and the percentage of errors in performing specific tasks. Ninety-four cirrhotic patients (aged 58 +/- 9 years) without overt hepatic encephalopathy and 80 controls (aged 53 +/- 15 years) were consecutively enrolled. The median follow-up in cirrhotic patients was 426 days (lower quartile = 213 days; upper quartile = 718 days). Results of the NCT, Scan test, and Choice2 test were significantly worse in cirrhotic patients, whereas Choice1 did not differ significantly from the controls. In cirrhotic patients, the prevalence of altered psychometric tests was 21% (CI95% = 14%-31%) by NCT, 23% (CI95% =15%-33%) by Scan test, and 20% (CI95% =16%-30%) by Choice2 test. The alterations of NCT, Scan, and Choice2 were found to be related to the severity of liver disease, independently of its etiology. Increased risk of death was found to be associated with altered Scan test (hazard ratio = 2.4; CI95% =1. 1-5.3), or altered Choice2 test (hazard ratio = 2.8; CI95% = 1.2-6. 3). Multivariate regression showed that Scan and Choice2 tests had prognostic value on survival, in addition to Child-Pugh classes in the first year of follow-up.