In order to estimate the consistency of variables measured in different test situations classically used in psychopharmacology, a set of hybrid mice was confronted with five testing situations: responses to a novel object introduced in a familiar environment, responses to novel places, behavior in a light/dark choice situation, on the holeboard, and in an elevated plus maze. A principal component analysis was performed using two variables per device as active variables and 26 others as supplementary variables. The first factor was clearly due to opposition between high and low levels of neophobia. Only variables from the first two tests were strongly correlated to this factor. Variables from the holeboard and from the plus maze were highly correlated one another and with the second factor, which grouped locomotion and exploration criteria. The light/dark choice test was intermediate and seemed to be moderately related to both locomotion and neophobia. These results point out the difficulty in comparing different test devices from a psychological point of view.