The Morris water maze task is a widely used tool to assess hippocampus-dependent learning and memory in rodents. Performance depends upon several factors including not only the traits of the experimental animals, but also apparatus and protocol characteristics. The present study aimed at investigating the effect of maze diameter on acquisition and probe trial performance in three commonly used strains: C57Bl/6, BALB/c, and 129/SvEvBrd mice. Three maze diameters (150, 120, and 75cm) were used under identical protocol and testing conditions. Downscaling maze dimensions, hence reducing difficulty and stress levels, did not allow BALB/c mice, commonly known as poor learners, to acquire this visual-spatial learning task. C57Bl/6 mice performed satisfactory in all three maze settings, with superior probe trial performance in the 120-cm-diameter setting. Further downscaling of maze dimensions might even render this task too simple for this strain. If the 129S5/SvEvBrd background strain is preferred, testing of visual-spatial learning abilities should be performed in a small sized MWM pool, as this strain performed only adequately in the smallest maze setting. Attention is drawn to the importance of supplying a detailed description 129 substrain nomenclature in future studies. Generalization of observations from one strain to another and from data obtained with a specific strain and maze diameter to other maze dimensions should be dealt with very carefully. The present study emphasizes the importance of a well-substantiated choice of background strain and water maze characteristics when researchers plan to investigate visual-spatial learning and memory in a chemically/lesion-induced or targeted mutagenesis model.