Patterns of spike activity were measured in the pericruciate cortex of conscious cats before and after development of a Pavlovian conditioned eye blink response. Unit activity was tested with presentations of a click conditioned stimulus (CS) and a hiss discriminative stimulus (DS) of similar intensity to the click. Unit discharge in response to the CS increased after conditioning, but not after backward conditioning when conditioned reflexes (CRs) were not performed. Rates of spontaneous, baseline discharge were not increased after conditioning with respect to rates of discharge measured in the naive state. It appeared that an increase in the ratio of CS-elicited discharge to background activity, together with an increase in the number of units responding to the CS after conditioning, supported discrimination of the CS from the DS and performance of the conditioned blink response. This is the first detailed characterization of patterns of a rapidly conditioned Pavlovian response. Activation of units by the CS preceded the onset of the CR, supporting the hypothesis that the activity played a role in initiating the conditioned eye blink movement. Extinction with retention of performance of the CR was associated with perseverance of the increased unit discharge in response to the CS. Extinction with substantially reduced performance of the CR was associated with diminution of the unit response to the CS below levels found with conditioning. Averages of patterns of spike activity elicited by the CS after conditioning showed components of discharge with onsets of 8–40 msec (alpha 1), 40–72 msec (alpha 2), 72–112 msec (beta), and greater than 112 msec (gamma), corresponding to each of four separate excitatory EMG components of the compound blink CR. Each component increased in magnitude after conditioning, relative to levels found in the naive state. The finding that long- as well as short-latency components of unit activation increased after conditioning supported the hypothesis that generation of both long- and short-latency blink CRs in normal animals may depend significantly on neural circuitry and mechanisms within the motor cortex.