GTP, when added to a rat liver microsomal fraction that had previously been allowed to accumulate Ca2+, causes a slow release of Ca2+, which is greatly enhanced by addition of inositol trisphosphate (IP3). The Ca2+ release caused by IP3 under these conditions is very much greater than that observed in the absence of GTP. The effect of GTP is dependent on the presence of polyethylene glycol in the incubation medium and is not due to inhibition of the Ca2+-accumulation system. The response to GTP is time-dependent, particularly at low (4 microM) GTP concentrations, and cannot be mimicked by ATP, ITP, CTP, UTP and GDP. Studies with [gamma-32P]GTP show that, during incubation with microsomal fractions, the terminal phosphate of GTP is transferred to two protein species, of Mr 38 000 and 17 000. These protein phosphorylations are still present when an excess of unlabelled ATP is included in the incubation mixture, but appear to be unaffected by the presence or absence of IP3 and polyethylene glycol. As a working hypothesis, it is suggested that a protein, phosphorylated by GTP, has to bind to the microsomal membranes before IP3 can stimulate Ca2+ release, and that, in vitro, the binding of this protein is favoured by the presence of polyethylene glycol.