NK cells are large granular lymphocytes that are an important component of the innate immune system. Surface density expression of CD16 and CD56 can be used to classify functionally and developmentally distinct NK cell subsets. This notion has more recently been confirmed by the identification of unique cytokine receptor expression patterns within each NK cell subset. There are substantial data to suggest that NK cells are derived from proliferating bone marrow precursors. Murine studies have also shown that an intact marrow environment is necessary for the development of NK cells from their progenitor populations. Culture of bone marrow cells in IL-2-containing medium gives rise to effectors that are essentially indistinguishable from mature NK cells, and it has also been shown that NK cells can be generated in vitro by culture of CD34(+) bone marrow cells with IL-2. Several lines of evidence suggest that NK cells and T cells may share a common precursor. T cells and NK cells can both be generated from immature thymocytes in vitro, and both CD3(+) lymphocytes with functional TCR and CD3(-)CD16(+) NK cells can be generated from fetal liver cells under the appropriate culture conditions. One might therefore postulate the existence of a common NK/T-cell progenitor within the fetal liver that possesses the ability to differentiate into T lymphocytes under thymic influences. Alternatively, in the absence of these signals, or perhaps in response to other stimuli (i.e., IL-2), the NK/T-cell progenitor differentiates into cytolytic NK effectors. Although IL-2 has been found to be critical to the development of NK cells in vitro, it is important to note that NK cells are present in IL-2 knockout mice. Therefore, other cytokines that bind to components of the IL-2R may be important to the development of NK cells in vivo.