The effects of two training and two surgical procedures on the recovery of a visual pattern discrimination were compared in rats with visual cortex damage (areas 17, 18, and 18a). Normal control subjects and rats with serial or simultaneous visual cortex lesions were first trained on a brightness (black vs. white) discrimination task and then on a pattern (horizontal vs. vertical stripes) discrimination task by either classical or fading procedures. The rats' performance in the preoperative, interoperative, and postoperative stages of training were compared. Rats with serial lesions showed postoperative savings in the brightness task and relearned the pattern discrimination task when trained by either the fading or the classical method. Serially lesioned subjects presented fewer errors to criterion when trained by the fading procedure than when trained by the classical procedure. Rats with simultaneous lesions showed no savings on the brightness discrimination task and all of the subjects in this group relearned the pattern discrimination task when trained by the fading procedure. Simultaneously lesioned subjects trained to a maximum of 750 trials using the classical method did not relearn the pattern discrimination task, except for one subject. Interaction between the surgical and training variables is discussed.It has long been known that, after simultaneous bilateral removal of the visual cortex (areas 17, 18, and 18a, as defined by Krieg, 1946aKrieg, , 1946b or all the posterior half of the cortex, rats are able to relearn a black vs. white (B-W) brightness discrimination task