Three rhesus monkeys self-administered fixed doses (4, 8 or 16 mg/kg) of an orally-delivered phencyclidine solution (0.25 mg/ml) under a one response, fixed-ratio (FR 1) schedule. They self-administered a saccharin 0.05% (wt/vol) solution, under an FR 8 schedule, 40 minutes after phencyclidine access began. After behavior stabilized over at least five daily sessions at each dose, water was substituted for phencyclidine for four sessions, and then phencyclidine was reinstated at the same dose, and behavior was allowed to stabilize. A comparison of saccharin-maintained behavior before and after the four sessions of water substitution yielded a two-fold shift in the dose effect curve indicating a rapid loss of tolerance to phencyclidine. Phencyclidine produced a biphasic effect on saccharin-maintained behavior: response rates increased at the low dose and decreased at the high dose. Subsequently, the FR requirement for saccharin was increased to 16 and the number of sessions of water substitution was varied: 1, 4, 8 or 16. The results indicated that tolerance was lost after four days of phencyclidine absence but not one, and complete tolerance was acquired within four to five days. This procedure produced a reliable measure of tolerance under limited access conditions whereby orally-delivered phencyclidine was also demonstrated to be functioning as a reinforcer.