Experiments were performed to investigate the effects of microinjections of the 5-HT(1A) agonist 8- hydroxy-2-(di-n-propylamino) tetralin (8-OH-DPAT), the non-selective 5-HT(1) agonist 5-carboxamidotryptamine (5-CT) and the GABA(A) agonist muscimol into the midbrain raphe nuclei, on behaviour maintained by a differential reinforcement of low rate (DRL) 20 schedule of food reinforcement. Following acquisition of responding under the DRL20 schedule, in which responses were reinforced only if they occurred at least 20s after the previous response, two groups of rats were prepared with a stainless steel guide cannula aimed at either the dorsal raphe nucleus or the median raphe nucleus. Injections of 8-OH-DPAT and 5-CT into the median raphe dose-dependently increased responding and reduced the number of reinforcers earned, leading to a net reduction in response efficiency from 45% to approximately 27% (5µg 8-OH-DPAT) and 22% (375ng 5-CT). Both drugs appeared to shift the frequency distribution of inter-response times (IRTs) towards shorter IRTs, and lowered the mean IRT. These effects were not observed after dorsal raphe injections of either drug. This pattern of results, together with results obtained in other paradigms measuring response inhibition, suggests that suppression of the activity of median raphe 5-HT neurones induces deficits in the ability to withhold responding. Muscimol increased responding, decreased the number of reinforcers earned and reduced response efficiency after both dorsal raphe and median raphe injections. These effects were more pronounced following median raphe injections, and were of considerably greater magnitude than those observed following treatment with the 5-HT agonists. Muscimol injected into the median raphe lowered the mean IRT, and increased the frequency of short duration IRTs. Thus, stimulation of GABA(A) receptors within the median raphe induces a pattern of behavioural disruption in the DRL task, that is more severe than that resulting from selective inhibition of 5-HT neural activity. The effects of muscimol probably arise from a general behavioural activation, rather than a specific deficit in the ability to withhold responding.