1. Intracellular recording techniques were used to study the effects of repetitive stimulation on monosynaptically activated inhibitory postsynaptic currents (IPSCs) in rat hippocampal slices. This was achieved by stimulation in stratum radiatum close to a recorded CA1 pyramidal neurone after pharmacological blockade of excitatory synaptic responses, using a combination of the N-methyl-D-aspartate (NMDA) and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptor antagonists D-2-amino-5-phosphonopentanoate (AP5; 0.04-0.1 mM) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; 0.02-0.04 mM), respectively. 2. Fixed-intensity stimulation at frequencies of less than 0.1 Hz evoked biphasic IPSCs of constant amplitude and waveform. In contrast, when two shocks (paired pulse) or longer trains of ten or more stimuli (i.e. tetani) were delivered at frequencies of between 0.2 and 20 Hz there was marked depression of both phases of every IPSC (by 60-100%) relative to the first or 'priming' IPSC evoked. 3. The gamma-aminobutyric acid (GABA)B receptor antagonists phaclofen (0.4-2 mM), 2-hydroxy-saclofen (0.02-0.4 mM) and 3-aminopropyl(diethoxymethyl)phosphinic acid (CGP 35348; 0.01-1 mM) reduced or abolished, in a concentration-dependent and reversible manner, both the late phase of the IPSC (IPSCB) and paired-pulse depression of the early phase of the IPSC (IPSCA). Expressed in terms of IC50 values, all three antagonists were 5-10 times more potent at blocking IPSCB than paired-pulse depression. 4. Paired-pulse depression, at 5 and 10 Hz, has been shown to be mediated by GABA acting on presynaptic GABAB receptors (i.e. GABAB autoreceptors). We now show that GABAB receptor antagonists reverse paired-pulse depression over the entire range of frequencies (0.1-50 Hz) that it occurs. 5. GABAB receptor antagonists reversed substantially the depression of IPSCs during tetani delivered at 5 or 10 Hz. However at 20 Hz, GABAB receptor antagonists appeared to be less effective. At 100 Hz they appeared to be ineffective at reversing the depression of IPSCA; since the antagonists block IPSCB the net effect was to reduce the level of outward current. 6. At frequencies of 20 Hz or more, there was also the appearance of a slow inward current which increased in size in proportion to the frequency and number of shocks in the tetanus. This current (termed here IPSCI) was more pronounced at hyperpolarized membrane potentials and was blocked by picrotoxin (0.1 mM) or bicuculline (0.05 mM). 7. 'Priming' is considered to represent a more physiological pattern of activity than a tetanus.(ABSTRACT TRUNCATED AT 400 WORDS)