Cell pairs in the auditory cortex of the cat were recorded with one micro-electrode and a computer system which could distinguish between their spikes through a self-learning pattern recognition algorithm. Synaptic connections were derived from the cross-correlation function between the two spike-trains. Each pair was assigned to one of three classes: No connection (NC), excitatory connection (EXC) and common input from a third neuron (CI). Inhibitory influences could not be revealed due to their weak expression in cross-correlograms. We found 22% of cell-pairs having excitatory connections, and 52% which shared a common input. 36% of directly neighbouring neurons show no correlation between their spike-trains. 10% were of mixed EXC&CI type