•  Previous studies indicate that cholinergic neuromodulation is required for cognitive processes and for gamma oscillatory activity in neocortical networks in vivo. The cholinergic agonist carbachol (CCh) induces gamma oscillations in vitro, via mechanisms that may be shared with those mediating in vivo gamma oscillations. •  Here, we studied the effects of CCh on cortical circuit components thought to be critical for gamma oscillations, and found that CCh stimulated firing of pyramidal cells (PCs) and increased excitatory synaptic input onto fast‐spiking interneurons (FSNs). •  CCh also modulated synaptic transmission between FSNs and PCs, decreasing synaptic depression during repetitive presynaptic firing, while simultaneously reducing the unitary synaptic currents. •  CCh increased the probability of neuron firing per oscillation cycle when PCs and FSNs fired in response to oscillatory input at gamma frequency. •  Combined, these effects of CCh may help explain the contribution of cholinergic modulation to gamma oscillations. Abstract  Cholinergic neuromodulation in neocortical networks is required for gamma oscillatory activity associated with working memory and other cognitive processes. Importantly, the cholinergic agonist carbachol (CCh) induces gamma oscillations in vitro, via mechanisms that may be shared with in vivo gamma oscillations and that are consistent with the pyramidal interneuron network gamma (PING) model. In PING oscillations, pyramidal cells (PCs), driven by asynchronous excitatory input, recruit parvalbumin‐positive fast‐spiking interneurons (FSNs), which then synchronize the PCs via feedback inhibition. Whereas the PING model is favoured by current data, how cholinergic neuromodulation contributes to gamma oscillation production is poorly understood. We thus studied the effects of cholinergic modulation on circuit components of the PING model in mouse medial prefrontal cortex (mPFC) brain slices. CCh depolarized and evoked action potential firing in a fraction of PCs and increased excitatory synaptic input onto FSNs. In synaptically connected pairs, CCh reduced the short‐term depression at FSN–PC and PC–FSN synapses, equalizing synaptic strength during repetitive presynaptic firing while simultaneously increasing the failure probability. Interestingly, when PCs or FSNs fired in response to gamma frequency oscillatory inputs, CCh increased the firing probability per cycle. Combined with the equalization of synaptic strength, an increase by CCh in the fraction of neurons recruited per oscillation cycle may support oscillatory synchrony of similar strength during relatively long oscillation episodes such as those observed during working memory tasks, suggesting a significant functional impact of cholinergic modulation of mPFC circuit components crucial for the PING model.