Fructose feeding increases hepatic de novo lipogenesis (DNL) and is associated with nonalcoholic fatty liver disease. Little is known, however, about individual variation in susceptibility to fructose stimulation of DNL. In this three-period crossover study, 17 healthy male subjects were enrolled to evaluate the within- and between-subject variability of acute fructose feeding on hepatic fractional DNL. During each assessment, [1-¹³C₁]acetate was infused to measure DNL in the fasting state and during fructose feeding. Subjects randomly received a high dose of fructose (10 mg·kg fat-free mass⁻¹·min⁻¹) on two occasions and a low dose (5 mg·kg fat-free mass⁻¹·min⁻¹) on another. Fructose solutions were administered orally every 30 min for 9.5 h. Ten subjects completed all three study periods. DNL was assessed as the fractional contribution of newly synthesized palmitate into very-low-density lipoprotein triglycerides using mass isotopomer distribution analysis. Mean fasting DNL was 5.3 ± 2.8%, with significant within- and between-subject variability. DNL increased dose dependently during fructose feeding to 15 ± 2% for low- and 29 ± 2% for high-dose fructose. The DNL response to high-dose fructose was very reproducible within an individual (r = 0.93, P < 0.001) and independent of fasting DNL. However, it was variable between individuals and significantly correlated to influx of unlabeled acetyl-CoA (r = 0.7, P < 0.001). Unlike fasting DNL, fructose-stimulated DNL is a robust and reproducible measure of hepatic lipogenic activity for a given individual and may be a useful indicator of metabolic disease susceptibility and treatment response.