We are currently studying the mechanisms of operantly conditioned changes in the H-reflex in the rat. Primate data suggest that H-reflex decrease is due to a positive shift in motoneuron firing threshold and a small decrease in the monosynaptic excitatory postsynaptic potential (EPSP), and that increase might be due to change in group-I oligosynaptic (especially disynaptic) input. To further evaluate the possibility of conditioned change in oligosynaptic input, we compared the mean latency (i.e., the average latency of the entire H-reflex) and the duration of control (i.e., pre-conditioning) H-reflexes with those of H-reflexes after up-conditioning or down-conditioning. Up-conditioning was associated with small, statistically significant increases in H-reflex mean latency [+0.11+/-0.05 (+/-SE) ms] and duration (+0.32+/-0.16 ms). The mean latency of the H-reflex increase (i.e., the part added to the H-reflex by up-conditioning) was 0.28+/-0.14 (+/-SE) ms greater than that of the control H-reflex. Down-conditioning had no significant effect on mean latency or duration. While these results indicate that operant conditioning does not greatly change H-reflex mean latency or duration, the effects detected with up-conditioning are consistent with the hypothesis that decreased inhibition, or increased excitation, by homonymous and heteronymous group-I oligosynaptic input contributes to the H-reflex increase produced by up-conditioning. Several other mechanisms might also account for these small effects.