The two yeast pheromone receptors, the a and -factor receptors, share many functional similarities: both G protein-coupled receptors couple to the same downstream signal transduction pathway, and both receptors undergo feedback regulation involving increased phosphorylation on their C-terminal domains in response to ligand challenge. The present work, which focuses on the signaling mechanism controlling this feedback phosphorylation, indicates one striking difference. While the -factor-induced phosphorylation of the -factor receptor does not require activation of the downstream G protein-directed signaling pathway (B. Zanolari, S. Raths, B. Singer-Kruger, and H. Riezman, Cell 71:755-763, 1992), the a-factor-induced phosphorylation of the a-factor receptor (Ste3p) clearly does. Induced Ste3p phosphorylation was blocked in cells with disruptions of various components of the pheromone response pathway, indicating a requirement of pathway components extending from the G protein down through the mitogen-activated protein kinase (MAPK). Furthermore, Ste3p phosphorylation can be induced in the absence of the a-factor ligand when the signaling pathway is artificially activated, indicating that the liganded receptor is not required as a substrate for induced phosphorylation. While the activation of signaling is critical for the feedback phosphorylation of Ste3p, pheromone-induced gene transcription, one of the major outcomes of pheromone signaling, appears not to be required. This conclusion is indicated by three results. First, ste12 cells differ from cells with disruptions of the upstream signaling elements (e.g., ste4, ste20, ste5, ste11, ste7, or fus3 kss1 cells) in that they clearly retain some capacity for inducing Ste3p phosphorylation. Second, while activated alleles of STE11 and STE12 induce a strong transcriptional response, they fail to induce a-factor receptor phosphorylation. Third, blocking of new pheromone-induced protein synthesis with cycloheximide fails to block phosphorylation. These findings are discussed within the context of a recently proposed model for pheromone signaling (P. M. Pryciak and F. A. Huntress, Genes Dev. 12:2684-2697, 1998): a key step of this model is the activation of the MAPK Fus3p through the G-dependent relocalization of the Ste5p-MAPK cascade to the plasma membrane. Ste3p phosphorylation may involve activated MAPK Fus3p feeding back upon plasma membrane targets.