Chronic alcohol intake is associated with an increased incidence of certain neoplasms. Natural killer (NK) cells have been considered to be involved in control tumor development and growth. The goal of the present study was to contribute to a better understanding of the effects of ethanol (EtOH) per se on the NK-cell population. Both patients with chronic alcoholism without liver disease (AWLD) and subjects with alcohol-induced cirrhosis (ALC) were carefully selected for this study. Immunophenotypical and functional studies of peripheral blood (PB) NK-cells were performed during active EtOH intake and after 3 months of a withdrawal period. In the AWLD group a significant increase in number of NK-cells (CD3-/CD56+) (P < .05) associated with a parallel increase in NK-cell lytic activity (P < .01) was observed. In addition, the number of cytotoxic T cells displaying the CD3+/CD56+ phenotype as well as CD8-/CD57+ NK-cell subset was also increased (P < .01 and P < .001, respectively). By contrast, in ALC patients with active EtOH intake (ALCET group), although a significant increase in the number of NK PB lymphocytes was observed (P < .05), NK lytic activity was depressed (P < .05), suggesting the existence of a decreased lytic activity/NK-cell. After 3 months of EtOH withdrawal, PB mononuclear cells (PBMC) from the AWLD group patients still displayed an increased NK cytolytic activity; in addition, the number of PB NK-cells (CD3-/CD56+ and CD8-/CD57+) and CD3+/CD56+ PB T cells continued to be increased. Independently of the duration of withdrawal period, in ALC patients EtOH withdrawal was followed by a slight decrease in the NK lytic activity of PBMC with respect to the values in active alcoholism phase; slight differences observed in the NK lytic activity in ALC patients who quit drinking could be related to the tendency to decrease of the number of NK-cells toward normal values. Furthermore, although an increase in NK cytotoxic activity after stimulation of PBMC with interleukin-2 (IL-2) was observed in ALC, it did not reach the levels observed in healthy subjects. Overall, our results show that the behavior of PB NK-cell population in chronic alcoholism is different according to both the moment of EtOH consumption and the existence or not of ALC. Alcohol by itself induced an increase in the number and lytic activity of NK-cells. By contrast, the NK cytolytic activity is constantly depressed in the stage of alcoholic cirrhosis, supporting the notion that immunosurveillance mechanisms may be affected in these patients.