Central thalamus has extensive connections with basal ganglia and frontal cortex that are thought to play a critical role in sensory-guided goal-directed behavior. Central thalamic activity is influenced by cholinergic projections from mesopontine nuclei. To elucidate this function we trained rats to respond to lights in a reaction time (RT) task and compared effects of muscarinic (2.4, 7.3, 22 nmol scopolamine) and nicotinic (5.4, 16, 49, 98 nmol mecamylamine) antagonists with the GABA(A) agonist muscimol (0.1, 0.3, 1.0 nmol) in central thalamus. We compared this with subcutaneous (systemic) effects of mecamylamine (3.2, 9.7, 29 micromol/kg) and scopolamine (0.03, 0.09, 0.26 micromol/kg). Subcutaneous scopolamine increased omissions (failure to respond within a 3-s response window) at the highest dose tested. Subcutaneous mecamylamine increased omissions at the highest dose tested while impairing RT and per cent correct at lower doses. Intrathalamic injections of muscimol and mecamylamine decreased per cent correct at doses that did not affect omissions or RT. Intrathalamic scopolamine increased omissions and RT at doses that had little effect on per cent correct. Anatomical controls indicated that the effects of mecamylamine were localized in central thalamus and those of scopolamine were not. Drug effects did not interact with attention-demanding manipulations of stimulus duration, proximity of stimulus and response locations, or stimulus array size. These results are consistent with the hypothesis that central thalamus mediates decisional processes linking sensory stimuli with actions, downstream from systems that detect sensory signals. They also provide evidence that this function is specifically influenced by nicotinic cholinergic receptors.