To remember what one did yesterday is an example of an everyday episodic memory task, in which a female advantage has sometimes been reported. Here, we quantify the impact of sex on episodic memory performance and investigate whether the magnitude of the sex difference is modified by study-, task-, and sample-specific moderators. Analyses were based on 617 studies conducted between 1973 and 2013 with 1,233,921 participants. A 5-level random-effects meta-analysis showed an overall female advantage in episodic memory (g = 0.19, 95% CI [0.17, 0.21]). The material to be remembered affected the magnitude of this advantage, with a female advantage for more verbal tasks, such as words, sentences, and prose (g = 0.28, 95% CI [0.25, 0.30]), nameable images (g = 0.16, 95% CI [0.11, 0.22]), and locations (g = 0.16, 95% CI [0.11, 0.21]), and a male advantage in more spatial tasks, such as abstract images (g = −0.20, 95% CI [−0.35, –0.05]) and routes (g = −0.24, 95% CI [−0.35, –0.12]). Furthermore, there was a female advantage for materials that cannot easily be placed along the verbal-spatial continuum, such as faces (g = 0.26, 95% CI [0.20, 0.33]), and odor, taste, and color (g = 0.37, 95% CI [0.18, 0.55]). These differences have remained stable since 1973. For verbal episodic memory tasks, differences were larger in Europe, North America, Oceania, and South America than in Asia, and smaller in childhood and old age than for other ages. Taken together, results suggest that men may use their spatial advantage in spatially demanding episodic memory tasks, whereas women do well in episodic memory tasks that are verbalizable and tasks that are neither verbal nor spatial, such as remembering faces and odors/tastes/colors.