The existence of sex differences in the standard rat and mouse models of learning and memory is a controversial and contested topic in the literature. The present meta-analysis of radial maze and water maze experiments was conducted to assess the reliablility and magnitude of sex effects in the standard rodent models of learning and memory. Data were culled from published and unpublished sources. Findings indicate large reliable male advantages for rats in radial maze and water maze protocols. Significant strain differences were also identified. In each paradigm, protocol variations were associated with differential sex effects. For the water maze, smaller male advantages were associated with pretraining regimens and for the radial maze, larger significant male advantages were observed in protocols that included unbaited arms (combined reference and working memory protocols). Mouse studies exhibited a different pattern of sex effects; small female advantages were evident in the water maze, but small male advantages were evident in the radial maze. Together these findings establish the reliability of male advantages in spatial working and reference memory for rats across strains, protocols, ages and rearing environments. The findings also support an important species dichotomy between rats and mice that should be considered when transitioning from rat to mouse models. In light of these results, the biological evidence supporting theoretical explanations of sex differences is reviewed and evaluated.