How adenosine leakage and tetanic release might affect long-term potentiation (LTP) was investigated by applying adenosine antagonists 8(p-sulfophenyl)theophylline (8SPT) or 8-cyclopentyl-3,7-dihydro-1,3-dipropyl-1H-purine-2,6-dione (DPCPX) to slices, while recording CA1 field EPSPs and population spikes. In the first series of experiments, we applied weak double tetani (at 100 Hz, for 1 s) that were subliminal for evoking LTP in initial control runs. In the presence of 8SPT--at concentrations (10-50 microM) which block both A1 and A2 receptors--the same tetani consistently evoked LTP of population spikes but not of excitatory postsynaptic potentials (EPSPs), whereas DPCPX (50 nM), which blocks only A1 receptors, facilitated LTP of both EPSPs and population spikes. These results are consistent with previous evidence that tetanic adenosine release on the one hand depresses LTP via A1 receptors but on the other facilitates LTP via A2 receptors. In a second set of experiments, 8SPT (50-100 microM) did not prevent the induction of LTP of both EPSPs and population spikes by stronger tetanic stimulation. Therefore A2 receptor activation is not essential for the induction of LTP when stronger tetani are applied. Overall, the main effect of endogenous adenosine release is to oppose LTP induction.