Pekin ducklings (Anas platyrhynckos) were exposed either to a white or harlequin duck model. When tested for their preferences with both models simultaneously present, the harlequin was more often preferred. If tested in the presence of another strange object (a stuffed barn owl, Tytus alba), the harlequin-trained ducklings more often deviated from chosing their training model than did the white-trained ducks, i.e., a reversal of effects. Apparently, the context of the test interacts with the characteristics of the model in a way that confounds predictions.