Thyrotropin-releasing hormone (TRH) has 3 effects on clonal strains of rat pituitary cells in culture (GH-cells). Two long-term effects of TRH on GH-cells, which are measurable after 3 h or longer, have been previously reported; these are an increase in prolactin synthesis and a decrease in growth hormone production. We report here that TRH also stimulates the rapid release of stored intracellular prolactin. We have investigated the role of cyclic AMP as a possible mediator of the effects of TRH on GH-cells. Cyclic AMP concentrations are higher in cells treated with TRH compared with paired controls; a maximum difference of greater than 150% of control values is detected at 15 min if the incubation is performed in serum-free medium in the presence of 1 mM theophylline. The concentration of TRH required to give half-maximum increases in both prolactin release and cyclic AMP accumulation is 0.3 nM; half-maximal increases in prolactin synthesis occur at 3 nM TRH. Exogenous cyclic AMP (1 mM) causes only a slight increase in prolactin release; 8-bromo-cyclic AMP and 8-methylthio-cyclic AMP (1 mM) do not cause significant release. Phosphodiesterase inhibitors (0.3 mM theophylline, 0.03 mM isobutyl-methylxanthine) increase prolactin release but their effects on hormone synthesis are more complicated. Isobutylmethylxanthine, 8-bromo-cyclic AMP and 8-methylthio-cyclic AMP (0.4 MM) increase prolactin synthesis, but do not significantly affect growth hormone synthesis. Theophylline increases the synthesis of both hormones. Dibutyryl cyclic AMP (0.5 mM or more) increases prolactin release and both growth hormone and prolactin synthesis, but equivalent amounts of sodium butyrate have the same effects. We conclude that in GH-cells under carefully defined experimental conditions: 1) TRH causes an increase in intracellular cyclic AMP concentrations; 2) the increase in endogenous cyclic AMP and the effects of phosphodiesterase inhibitors are consistent with a model with cyclic AMP as a mediator of the effects of TRH on prolactin release; however, they do not prove this model, because the interpretation of these results depends on assumptions which may not all be valid; and 3) none of the analogs of cyclic AMP or the phosphodiesterase inhibitors tested mimic the decrease in growth hormone production caused by TRH.