Objective Fructose has become a focus of concern. To assess the effect of fructose on glycemic control, we conducted a systematic review and meta‐analysis. Methods We searched MEDLINE, EMBASE, and Cochrane Library (up to Nov 5, 2012) for relevant controlled trials ≥7‐days. Two independent reviewers extracted data. Data were pooled using random effects models for glycated blood proteins (GBP) and fasting glucose (FG) and insulin (FI). Results Eligibility criteria were met by 47 isocaloric trials (n=863), in which fructose was exchanged isocalorically for other carbohydrate, and 12 hypercaloric trials (n=143), in which the diet was supplemented with excess energy from fructose. Fructose in isocaloric trials reduced GBP (standardized mean difference = − 0.29 [95% CI: −0.46, −0.11]), equivalent to a ~0.59% reduction in HbA1c, in people with and without diabetes. In contrast, fructose in hypercaloric trials increased FG (mean difference [MD] = 0.18 mmol/L [95% CI: 0.08, 0.29]) and FI (MD = 6.06 pmol/L [95% CI: 3.70, 8.31]). Limitations Most trials were small, short, and of poor quality. Conclusions Isocaloric exchange of fructose for other carbohydrate improves glycemic control in people with and without diabetes. Fructose providing excess energy, however, raises fasting glucose and insulin levels, an effect that may be more owing to excess energy than fructose. Funding : Canadian Institutes of Health Research. Grant Funding Source : CIHR