Perfusion of isolated rat hearts with 150 mM beta GPA led to the linear accumulation of intracellular P beta GPA (approx. 150 nmol/min per g (dry wt.)) after an initial lag period of 20 min. 2. This accumulation of intracellular P beta GPA was accompanied by a decrease in PCr (30%) and an increase in total phosphagen content (20%). These results show that PCr was not equally replaced by P beta GPA, but was degraded at the expense of beta GPA phosphorylation to produce a net increase in cardiac phosphagen content. Correspondingly, total phosphate (the sum of PCr, P beta GPA, Pi and ATP) was increased, indicating that there was no cellular necrosis and that the sarcolemma remained intact throughout the perfusion. 3. An increase in Pi and decrease in ATP also occurred concomitantly with P beta GPA accumulation, indicating that ATP synthesis was not keeping up with demand. This may be due to the gradual replacement of PCr by the less efficient phosphagen, P beta GPA, resulting in inadequate transduction of energy and hence an imbalance between energy demand and supply. However, the increased hyperosmolarity of the perfusate may be partly responsible for these effects on cardiac energy metabolism, as perfusion with 150 mM mannitol produced a similar decrease in ATP, but a smaller rise in Pi. 4. Perfusion with either 150 mM beta GPA or mannitol led to a significant intracellular alkalosis (max. pHi 7.3), which was reversed on returning to normal perfusate. In addition, both hyperosmolar perfusions led to a significant reduction in cardiac frequency (40 and 15%, respectively). However, only beta GPA caused significant negative inotropism. The time-courses for the changes in cardiac frequency and pHi did parallel the increase in P beta GPA. This suggests that both hyperosmolarity and the production of P beta GPA during beta GPA perfusions determine the degree of negative chronotropism, but that hyperosmolarity alone causes alkalosis and beta GPA phosphorylation, a decrease in developed tension. 5. When hearts, acutely loaded with P beta GPA were perfused with control medium, the levels of ATP, PCr and P beta GPA stabilised to produce a new steady state. There was no decrease in P beta GPA concentration during this procedure, implying that beta GPA efflux was negligible.