In contrast to familiar role for Ras in proliferation, we and others previously suggested that Ras also mediates hypertrophy, the increase in cell mass characteristic of post-natal ventricular muscle. We showed that activated (G12R) and dominant-negative (S17N) Ha-Ras regulate "constitutive" and growth factor-responsive genes equivalently, in both cardiac myocytes and non-cardiac, Mv1Lu cells. Here, we attempt to delineate pathways by which Ras exerts this global effect. The E63K mutation, which impairs binding of guanine nucleotide releasing factor to Ras, alleviated suppression by S17N, consistent with sequestration of exchange factors as the mechanism for inhibition. To compare potential Ras effector proteins, we first engineered G12R/D38N, to abolish binding of Raf and phosphatidylinositol-3-kinase and established that this site was indispensable for augmenting gene expression. To distinguish between inhibition of Ras by Ras GTPase-activating protein (GAP) versus a potential effector function of GAP, we tested the effector domain substitution P34R: this mutation, which abolishes GAP binding, enhanced Ras-dependent transcription in Mv1Lu cells, yet interfered with Ras-dependent expression in ventricular myocytes. To examine the dichotomous role of Ras-GAP predicted from these P34R results, we transfected both cell types with full-length GAP, the C-terminal catalytic domain (cGAP), or N-terminal Src homology domains (nGAP). In Mv1Lu cells, cGAP markedly inhibited both reporter genes, whereas GAP and nGAP had little effect. Antithetically, in ventricular myocytes, GAP and nGAP activated gene expression, whereas cGAP was ineffective. Thus, Ras activates gene expression through differing effectors contingent on cell type, and an effector-like function of GAP predominates in ventricular muscle.