We address the mechanism of hybrid resistance (HR) in vitro using NK effector cells and target lymphoblasts from H-2^b, H-2^d, and H-2^b/d mice. The 5E6 (Ly49C)⁺ subset of F1 NK cells lyse BALB/c (H-2^d) but not B6 (H-2^b) targets unless either anti-5E6 or anti-H-2K^b MAbs are present. H-2D^d transgenic B6 (D8) targets are not susceptible to F1 Ly49A⁺ effectors. Furthermore, 5E6⁺ Ly49A⁺ F1 effectors lyse B6 and BALB/c targets only in the presence of anti-5E6 and anti-Ly49A MAbs, respectively. Thus, recognition of H-2K^b by 5E6 and H-2D^d by Ly49A transduce independent inhibitory signals. Moreover, anti-5E6 MAbs enable 5E6⁺ BALB/c NK cells to lyse (BALB/c × B6)F1 targets. These data support the "missing self" and not the "hemopoietic histocompatibility antigen" hypothesis for HR. In addition, 5E6⁺ NK cells from BALB/c and BALB.B, but not B6 or (BALB/c × B6)F1, mice receive negative signals from both H-2^d and K^b class I antigens. Thus, allelic differences in 5E6 (C57BL versus BALB) may regulate recognition events by NK cells.