Considerable evidence indicates that males navigate large-scale space better than females, and some have previously attributed this difference to a greater ability of males to select or use an allocentric (cognitive mapping) navigational strategy. We directly tested this proposal by having males and females navigate in an "ambiguous" virtual Morris water maze environment that permitted participants to choose and use either an allocentric or an egocentric strategy. A novel probe trial at the end of training revealed which strategy each participant had been using and showed that the strategy selected by the greatest number of males and females was allocentric, and that this bias was even greater for females. Traditional measures of navigational performance (distance, latency, probe dwell time) indicated that overall, males were more efficient than females. However, this gender difference was not related to strategy choice: males were better than females regardless of strategy, though the difference was significant only in those navigating allocentrically. These data indicate that while males may navigate allocentrically more efficiently than females, this does not account for the male advantage in navigation. The data also indicate that under specific circumstances, females may also prefer and use an allocentric strategy to navigate. These findings have implications for theories regarding the differential use of the hippocampus by men and women.