We propose possible mechanisms for the effects of neuromodulators on the perception of time on the basis of our hypothesis that estimation of intervals between sensory stimuli depends on the beat frequency of a "internal clock," which is inversely proportional to the latency of reentering excitation of the neocortex. These effects are based on the modification of the efficacy of excitatory inputs to the cortical neurons, as well as inputs from the cortex to the striatum, which are necessary for the disinhibition of thalamic neurons by the basal ganglia and subsequent excitation of the cortex. The character of influence is determined by the con centration of neuromodulators, the types of receptors activated by them on neurons of the striatum and cor tex, and the modulation rules. According to the proposed mechanism, an increase in the frequency of the "internal clock" and overestimation of the duration of intervals may result from treatment with dopaminergic drugs, agonists of dopamine D1 and D2 receptors, and opioids and cannabinoids, which promote an increase in the dopamine concentration, as well as antagonists of adenosine A1 and A2A and muscarinic M2 recep tors, whose activation facilitates disinhibition of the thalamus by the basal ganglia. Antagonists of A1, M2, and D2 receptors, which prevent depression of excitatory inputs to neocortical pyramidal cells, also can increase the frequency of the "internal clock." The activation of a large number of D2 receptors on the cor tical pyramidal cells, which results from a considerable increase in dopamine concentration, like activation of cannabinoid CB1 receptors, should promote a decrease in the frequency of the "internal clock" and the underestimation of interval duration. The activation of D2 and M2 receptors on the GABAergic interneurons of the cortex under conditions of strong inhibition of pyramidal neurons may increase the beat frequency of the "internal clock." The proposed mechanism helps to understand the causes of errors in time perception in neurological diseases and to explain the discrepancies in the results of studies on the effects of neuromodula tors on the estimation of time. The hypothesis may be experimentally examined by treatment of the striatum or neocortex with agonists and antagonists of various types of receptors and measurement of the drug induced changes in the interval between the first and second peaks in the distribution of latencies of responses of cortical neurons to sensory stimulus.