Ontogenic development and the lymphokine responsiveness of human NK cell activity against K562 target cells in peripheral blood lymphocytes were evaluated in fetuses, premature infants, and term neonates by using a 4-hr 51Cr-release assay. Basal NK activity and NK boosting by lymphokines were comparatively assayed after an 18-hr incubation with medium alone, recombinant human IFN-gamma (1000 U/ml), and recombinant human IL 2 (25 U/ml), respectively. Lymphocytes from 20-wk-old fetuses lacked NK cell activity even after the pretreatment with IFN-gamma. Low, but significant levels of NK activity and NK boosting by IFN-gamma were observed in premature infants after 27 wk of gestation, with a progressive intrauterine maturation of these activities. Both basal NK activity and NK boosting by IFN-gamma in term neonates were still lower than those of adult controls. The grade of NK boosting by IFN-gamma appeared to depend on the development of basal NK activity. Contrary to IFN-gamma, IL 2 could induce marked NK activity even in 20-wk-old fetuses who lacked both basal and IFN-gamma inducible NK activities. NK boosting by IL 2 was much more efficient than that by IFN-gamma at any period of human life. The facts that IL 2-induced NK boosting could occur without any appreciable production of IFN-gamma in neonatal lymphocytes, and that ample neutralizing doses of anti-IFN-gamma antibody hardly suppressed IL 2-mediated NK boosting even in adult lymphocytes, indicated that the effect of IL 2 on NK boosting might be independent of IFN-gamma production. On the basis of the ontogenic differences in the development of the lymphokine responsiveness of NK cell activity and on the different NK boosting mechanisms of these lymphokines it was suggested that so-called human "pre-NK cells" might be divided into IFN-gamma sensitive and IL 2-sensitive cells. Whether these cell populations belong to different cell lineages or different maturation stages of the same cell line, however, remains unsettled.