We have reported earlier the isolation of two recessive, serum- and anchorage-dependent revertants from an NIH 3T3 line which had been transformed with multiple copies of a c-H-ras oncogene. In both revertants the oncogene was fully expressed and fusion of either revertant with normal (untransformed) cells or of the two revertants with one another resulted in transformed progeny. These, and other data indicate that the transforming activity of the c-H-ras oncogene is impaired in the two revertants, in consequence of defects in distinct genes needed to mediate this activity. Here, we describe some of the biochemical features of the revertants. In both of these (as in the transformed line) the bulk of the ras-p21 protein was found in the membrane fraction. This suggests proper posttranslational processing. Furthermore, no difference was detected either in the ras-p21 protein GTPase stimulating activity of GAP or in the extent of GAP-tyrosine phosphorylation among growing cultures of the two revertants, the transformed line and the parental NIH 3T3 line. The level of glucose transporter mRNA was severalfold higher in the transformed line than in the NIH 3T3 line. In the two revertants, however, the level was as low as that in the NIH 3T3 line. This indicates that the reversion impaired the effect of the c-H-ras oncogene on transcription. The raf oncogene (proposed to increase transcription factor activity) could retransform both revertants. Moreover, as revealed in experiments with growing cultures, neither transformation by the c-H-ras oncogene nor reversion from the transformed state altered the electrophoretic mobility of the raf protein or the level of its actin kinase activity. These results suggest that transformation by the c-H-ras oncogene is not mediated by the activation of raf protein kinase. The tyrosine phosphorylation of the p34cdc2 protein kinase (a cell cycle regulatory enzyme) was severalfold higher in the transformed line than in the NIH 3T3 line. The level of p34cdc2 protein kinase phosphorylation was as high in the R260 revertant as in the transformed line and as low in the R116 revertant as in the NIH 3T3 line. We are attempting to identify the defective mediator genes impairing the transforming activity of the c-H-ras oncogene in the two revertants.