In Xenopus oocytes injected with total rat pituitary GH3 cell RNA, thyrotropin‐releasing hormone (TRH) causes the activation of the inositol lipid transduction pathway and the induction of chloride conductance via calcium‐activated channels (Oron et al., 1987, Mol. Endocrinol., 1:918–925). This response exhibits characteristic prolonged latency (Oron et al., 1988, Proc. Natl. Acad. Sci. U.S.A., 85:3820–3824; Lipinsky et al., 1993, Pflugers Arch., 425:140–149). We examined the role of agonist diffusion in the extracellular medium in the generation of latency and the detemination of response amplitude. An increase in the viscosity of the medium markedly prolonged the latency and decreased the amplitude of the response. Moreover, an increase in the viscosity of the medium in the immediate vicinity of the oocyte had a major effect on both the latency and the amplitude of the response, which appeared to be a result of desensitization rather than restricted diffusion of chloride to the medium. Extrapolation to [TRH] ∞ yielded a diffusion‐dependent latency value of 0 and a diffusion‐independent latency value of 4 seconds. In low viscosity medium, at all TRH concentrations, diffusion contributed less than 2% to the latency of the response. This implied that events distal to ligand binding are responsible for a major part of latency. Analysis of the dependence of latency and amplitude of the response on [TRH] yielded Hill coefficients markedly smaller than unity, suggesting postreceptor negative modulation of the response. Preincubation of cells with a specific inhibitor of protein kinase C, chelerythrine, increased the Hill coefficients to unity and changed the shape of the Hill plot of response amplitudes. Our results suggest that at low agonist concentrations, even in a low viscosity medium, the prolonged latency allows negative effects on both latency and amplitude by a simultaneous activation of a protein kinase C. © 1995 Wiley‐Liss, Inc.