The object recognition test is now among the most commonly used behavioral tests for mice. A mouse is presented with two similar objects during the first session, and then one of the two objects is replaced by a new object during a second session. The amount of time taken to explore the new object provides an index of recognition memory. As more groups have used the protocol, the variability of the procedures used in the object recognition test has increased steadily. This protocol provides a necessary standardization of the procedure. This protocol reduces inter-individual variability with the use of a selection criterion based on a minimal time of exploration for both objects during each session. In this protocol, we describe the three most commonly used variants, containing long (3 d), short (1 d) or no habituation phases. Thus, with a short intersession interval (e.g., 6 h), this procedure can be performed in 4, 2 or 1 d, respectively, according to the duration of the habituation phase. This protocol should allow for the comparison of results from different studies, while permitting adaption of the protocol to the constraints of the experimenter.