Lean nonalcoholic fatty liver disease (NAFLD) is defined as NAFLD that develops in patients with a body mass index (BMI) less than 25 kg/m2. We investigated the differences between lean NAFLD and NAFLD in overweight and obese persons, factors associated with the severity of liver and cardiovascular disease, and the effects of visceral obesity. We performed a retrospective cohort study of 669 consecutive patients with biopsy-proven NAFLD seen at 3 liver centers in Italy. We collected anthropometric, clinical, and biochemical data, as well as information on carotid atherosclerosis (artery intima-media thickness and plaque), liver histology (nonalcoholic steatohepatitis [NASH] and fibrosis), insulin resistance, and diabetes. Overweight was defined as a BMI of 25 to 29.9 kg/m2, and obese was defined as a BMI of 30 kg/m2 or greater. Patients were assigned to groups based on waist circumference, a marker of visceral obesity (low: men, <94 cm, women <80 cm; medium: men, 94-102 cm, women 80-88 cm; or high: men >102 cm, women >88 cm). DNA samples were analyzed for the rs738409 C>G (I148M in PNPLA3), the rs58542926 C>T (E167K in TM6SF2), and single-nucleotide polymorphisms. Variables in men and women were analyzed using chi-squared analysis and the Mann-Whitney or Kruskal-Wallis tests. Multiple linear or logistic regression analyses were adjusted for all the variables of clinical relevance or statistically significant at univariate analyses. The primary outcome was the difference in liver and cardiovascular disease between lean NAFLD and NAFLD in overweight and obese persons. Secondary outcomes were effects of visceral obesity, based on waist circumference, on hepatic, vascular, and metabolic features. Significantly lower proportions of patients with lean NAFLD (143 patients; 43 women; mean age, 46 ± 13 y) had hypertension (P= .001), diabetes (P= .0001), and metabolic syndrome (P=.0001) than overweight or obese patients with NAFLD (526 patients; 149 women; mean age, 49 ± 12 y). Significantly lower proportions of patients with lean NAFLD had NASH (17% vs 40% of obese or overweight patients with NAFLD; P= .0001), fibrosis of F2 or higher (17% vs 42%; P=.0001), or carotid plaques (27% vs 39%; P= .03). Patients with lean NAFLD had significantly thinner carotid intima-media (0.74 ± 0.1 mm) than obese or overweight patients with NAFLD (0.84 ± 0.3 mm; P= .0001). There was no significant difference in the proportions of patients with rs738409 C>G in PNPLA3, but a significantly greater proportion of patients with lean NAFLD carried rs58542926 C>T in TM6SF2 (4%) than obese or overweight individuals with NAFLD (0.3%; P= .001). Of the 143 patients with lean NAFLD, 27 had grade 3 steatosis, 24 had a lobular inflammation score greater than 2, 10 had a ballooning score of 2, and 25 had a fibrosis score of 2 or higher. In patients with lean NAFLD, the only variable associated independently with NASH and a fibrosis score of 2 or higher was rs738409 C>G in PNPLA3. Patients with lean NAFLD and a medium waist circumference had a significantly higher risk of diabetes (odds ratio, 11; 95% confidence interval [CI], 1.2-106; P=.03) than overweight or obese patients with a similar waist circumference (odds ratio, 1.3; 95% CI, 0.4-4.2; P= .6). Lean and overweight or obese patients with high waist circumferences had significant increases in risk compared with patients with low and medium circumference and diabetes, hypertension, and fibrosis scores of 2 or higher. In a retrospective study of patients with lean NAFLD vs obese or overweight persons with NAFLD, we found 20% of patients with lean NAFLD to have NASH, fibrosis scores of 2 or higher, and carotid atherosclerosis. Lean patients with rs738409 C>G in PNPLA3 should be monitored for liver disease progression; studies including large series of patients with lean NAFLD will clarify the possible role of TM6SF2 polymorphisms.