Background: Conflicting evidence exists on the role of long-term fructose consumption on health. No systematic review has addressed the effect of isoenergetic fructose replacement of other sugars and its effect on glycated hemoglobin (HbA1c), fasting blood glucose, insulin, and triglycerides.Objective: The objective of this study was to review the evidence for a reduction in fasting glycemic and insulinemic markers after chronic, isoenergetic replacement of glucose or sucrose in foods or beverages by fructose. The target populations were persons without diabetes, those with impaired glucose tolerance, and those with type 2 diabetes.Design: We searched the Cochrane Library, MEDLINE, EMBASE, the WHO International Clinical Trials Registry Platform Search Portal, and clinicaltrials.gov The date of the last search was 26 April 2016. We included randomized controlled trials of isoenergetic replacement of glucose, sucrose, or both by fructose in adults or children with or without diabetes of ≥2 wk duration that measured fasting blood glucose. The main outcomes analyzed were fasting blood glucose and insulin as well as fasting triglycerides, blood lipoproteins, HbA1c, and body weight.Results: We included 14 comparison arms from 11 trials, including 277 patients. The studies varied in length from 2 to 10 wk (mean: 28 d) and included doses of fructose between 40 and 150 g/d (mean: 68 g/d). Fructose substitution in some subgroups resulted in significantly but only slightly lowered fasting blood glucose (-0.14 mmol/L; 95% CI: -0.24, -0.036 mmol/L), HbA1c [-10 g/L (95% CI: -12.90, -7.10 g/L; impaired glucose tolerance) and -6 g/L (95% CI: -8.47, -3.53 g/L; normoglycemia)], triglycerides (-0.08 mmol/L; 95% CI: -0.14, -0.02 mmol/L), and body weight (-1.40 kg; 95% CI: -2.07, -0.74 kg). There was no effect on fasting blood insulin or blood lipids.Conclusions: The evidence suggests that the substitution of fructose for glucose or sucrose in food or beverages may be of benefit to individuals, particularly those with impaired glucose tolerance or type 2 diabetes. However, additional high-quality studies in these populations are required.