The spontaneous object recognition (SOR) task is a versatile and widely used memory test that was only recently established in nonhuman primates (marmosets). Here, we extended these initial findings by assessing the performance of adult capuchin monkeys on the SOR task and three potentially intervening task parameters-object familiarization phase, retention delay and sex. In Experiment 1, after an initial 10-min familiarization period with two identical objects and a pre-established retention delay (0.5, 6 or 24h), the capuchins preferentially explored a new rather than the familiar object during a 10-min test trial, regardless of delay length. In Experiment 2, the capuchins were again exposed to two identical objects (but now for 10 or 20 min), then a 30-min retention delay and a 10-min test trial. An exploratory preference for the new over the familiar item was not affected by the length of the familiarization interval, possibly because overall exploration remained the same. However, the amount of initial object exploration was not related to task performance, and both males and females performed similarly on the SOR task with a 10-min familiarization, 30-min delay and 10-min test trial. Therefore, male and female capuchins recognize objects on the SOR task after both short and long delays, whereas a twofold increase in the familiarization phase does not affect task performance. The results also provide further support for the use of incidental learning paradigms to assess recognition memory in nonhuman primates.