Four pigeons performed on three types of schedules at short (i.e., 10, 30, or 60 s) interreinforcement intervals: (a) a delay‐dependent schedule where interreinforcement interval was held constant (i.e., increases in waiting time decreased food delay), (b) an interreinforcement‐interval‐dependent schedule where food delay was held constant (i.e., increases in waiting time increased interreinforcement interval), and (c) a both‐dependent schedule where increases in waiting time produced increases in interreinforcement interval but decreases in food delay. Waiting times were typically longer under the delay‐dependent schedules than under the interreinforcement‐interval‐dependent schedules. Those under both‐dependent schedules for 1 subject were intermediate between those under the other two schedule types, whereas for the other subjects waiting times under the both‐dependent procedure were similar either to those under the delay‐dependent schedule or to those under the interreinforcement‐interval‐dependent schedule, depending both on the subject and the interreinforcement interval. These results indicate that neither the interreinforcement interval nor food delay is the primary variable controlling waiting time, but rather that the two interact in a complex manner to