Researchers have examined if olfaction is a sensitive biomarker of anorexia nervosa, but considerable heterogeneity across studies makes it difficult to reach a consensus. This review and meta-analysis sought to clarify if olfaction is altered in individuals with anorexia nervosa and explore potential moderators of olfaction in this population. We performed quantitative and qualitative analyses of olfactory function in individuals with anorexia nervosa compared with healthy controls. A random effect model was used to estimate pooled effect sizes, and meta-regression was conducted to identify potential moderators. We found that individuals with anorexia nervosa had largely intact olfactory function compared with healthy controls in terms of threshold (g = −0.09, 95% confidence interval [CI] (−0.65,0.47), p = 0.757), identification (g = −0.06, 95% CI (−0.32,0.20), p = 0.642), and overall olfactory function (g = −0.47, 95% CI (−1.02,0.07), p = 0.090). Discrimination was different from control (g = −0.51, 95% CI (−0.97,-0.05), p = 0.029). However, after sensitivity analysis, the pooled effect size was nonsignificant in discrimination. Olfactory sensitivity covaried with anorexia nervosa severity, body mass index (BMI) positively moderated olfactory threshold score (β = 0.79, 95% CI (0.18,1.41), p = 0.020) in individuals with anorexia nervosa. Disease duration negatively moderated olfactory threshold score (β = −0.21, 95% CI (−0.40,-0.03), p = 0.034). The results suggest that olfaction is not a sensitive marker of anorexia nervosa diagnosis, but olfactory sensitivity may be a useful indicator of anorexia nervosa severity.