Six rhesus monkeys self-administered orally-delivered phencyclidine (PCP) and water under concurrent fixed-ratio (FR) 8 schedules. Liquids were available during three 6.5-h periods daily preceded by 1-h components when food was available under an FR 64 (lever press) schedule. After 10 days of stable behavior, water was substituted for PCP for 8 days. PCP was subsequently reinstated, and this PCP withdrawal sequence was repeated using different food FR values (64, 128, 256, 512 and 1024). Each time the food FR was changed behavior was allowed to stabilize for at least 10 days. Under all FR values food-maintained responding decreased markedly during PCP withdrawal, with a gradual recovery over the next 8 days. As the FR value increased from 64 to 1024 there was a parallel shift downward in food-maintained performance. When PCP was reinstated, food-reinforced responding generally returned to baseline rates during the first few days. In a second experiment monkeys were tested for PCP withdrawal effects under relatively food deprived or food satiated conditions under both an FR 512 and 1024 schedule of food delivery. The results showed that the decrease in food-maintained responding during withdrawal was inversely related to the total amount of food consumed during the control period. The results of these experiments indicate that manipulation of both the response requirements for food (FR) and the total amount of food available (food deprivation/satiation) alters the magnitude of response disruptions during PCP withdrawal.