Shapes of post-synaptic potentials (p.s.p.s) in cat motoneurones were compared with the time course of correlated changes in firing probability during repetitive firing. Excitatory and inhibitory post-synaptic potentials (e.p.s.p.s. and i.p.s.p.s) were evoked by brief triangular stretches of the triceps surae-plantaris muscles. Depolarizing current was injected through the recording micro-electrode to evoke repetitive firing and the post-stimulus time histogram of motoneurone spikes was obtained. E.p.s.p.s (n = 80) of different sizes (30-1040 microV) and rise times (1.1-8.2 ms) were investigated in fifty-nine motoneurones. The majority of the e.p.s.p.s were recorded in triceps surae-plantaris motoneurones with high levels of synaptic noise (estimated peak-to-peak fluctuations of 1.5-3.5 mV). This noise was generated by keeping the triceps surae-plantaris muscles stretched to a near maximal degree. The remaining e.p.s.p.s were recorded in motoneurones to other hind-limb muscles with a low level of synaptic noise. The height of the primary peak of the correlogram with respect to base-line firing rate increased in proportion to both amplitude and rising slope of the e.p.s.p.s. Using normalization procedures or using e.p.s.p.s of constant amplitude but different slopes and vice versa, the relative peak height increased with e.p.s.p. peak derivative with a slope of around 6/mV per millisecond and with e.p.s.p peak amplitude with a slope of about 1/mV. The shape of the correlogram (peak and trough) seemed well described by a linear combination of the shape of the e.p.s.p. derivative and that of the e.p.s.p. itself. The relative e.p.s.p. contribution (e.p.s.p.:e.p.s.p. derivative ratio) varied with e.p.s.p. amplitude and noise level, being largest (mostly 0.25-1.0) for small e.p.s.p.s (100-300 microV) in high levels of synaptic noise and smaller (0-0.25) for larger e.p.s.p.s and for e.p.s.p.s in a low noise background. In conformity with the above finding, a leaky integration of the correlograms gave time courses that in most cases closely resembled the shape of the e.p.s.p.s. The time constant of the leaky circuit (= inverse value of e.p.s.p.:e.p.s.p. derivative ratio) varied with e.p.s.p. amplitude and noise level in the same manner as obtained by direct fitting of e.p.s.p. and e.p.s.p. derivative shape to correlogram shape. Stretch-evoked i.p.s.p.s (n = 10) of small amplitude (90-360 microV as measured close to firing level) were investigated in pre-tibial flexor motoneurones with low levels of synaptic noise. These i.p.s.p.s generated correlogram troughs closely resembling, albeit somewhat wider than, the shape of the i.p.s.p. derivatives.(ABSTRACT TRUNCATED AT 400 WORDS)