The local application of electrical currents to the cortex is one of the most commonly used techniques to activate neurons, and this intracortical stimulation (ICS) could potentially lead to new types of neuroprosthetic devices that can be directly applied to the cortex. To identify whether ICS-activated circuits are physiological vs. profoundly artificial, it is necessary to record in vivo the responses of the same neuronal population to both natural sensory stimuli and artificial electric stimuli. However, few studies have extensively reported simultaneous electrophysiological recordings combined with ICS. Here, we evaluated the similarity between sound- and ICS-driven cortical response patterns in different cortical layers. In the mouse auditory cortex, we performed laminar recordings using 16-channel silicon electrodes and ICS using sharp glass-pipette electrodes containing biocytin for layer identification. In different cortical depths, short current pulses were delivered in vivo to mice under urethane anesthesia. For the recorded data, we mainly analyzed properties of local field potentials and current source densities (CSDs). We demonstrated that electrical stimulation evoked different excitation patterns according to the stimulated cortical layer; responses to electric stimuli in layer 4 were most likely to mimic acoustic responses. Next, we proposed a CSD-based stimulation method to artificially synthesize sound-driven responses, using an approximation method associated with a linear combination of CSD patterns electrically stimulated in the different cortical layers. The result indicates that synthesized responses were consistent with the canonical model of sound processing. Using these approaches, we provide a new technique in which natural sound-driven responses can be mimicked by well-designed computational stimulation pattern sequences in a layer-dependent manner. These findings may aid in the future development of an electrical stimulation methodology for a cortical prosthesis.