Objectives: To evaluate the association of predicted lean and fat mass on 10-year first and recurrent CVD incidence separately for men and women. Methods: Two prospective studies, ATTICA (2002-2012, n = 3042 subjects free-of-CVD, n = 1514 men (46 ± 13 years) and n = 1528 women (45 ± 14 years)) and GREECS (2004-2014, n = 2172 subjects with acute coronary syndrome (ACS), n = 1649 men (65 ± 13 years) and n = 523 women (62 ± 11 years)) were used. Lean mass index (LMI) and fat mass index (FMI) were created through total body lean and fat mass (indirectly calculated through population formulas based on body weight, height and waist circumference) divided by height squared. Follow-up was performed in n = 2020 of ATTICA (n = 317 first CVD events) and in n = 2172 patients of GREECS (n = 811 recurrent CVD events). Results: In ATTICA study, CVD rate from 1st to 3rd FMI tertile was 9.4%, 16.1% and 19.9% while in GREECS 36.2%, 37.0%, 38.3%. The LMI-related rates were 17.1%, 15.0% and 11.9% vs. 38.8%, 35.8% and 36.7%. Multiadjusted analysis revealed U-shape trend between LMI and CVD recurrence with 2nd LMI tertile having the best prognosis; this observation was more evident in women. In apparently healthy subjects, LMI-cardioprotective association was revealed only in 3rd tertile (HR = 0.91 95%CI (0.74, 0.95)); this was more evident in men. The FMI aggravating association (3rd tertile) was retained significant only in healthy women and ACS men. Conclusion: This work expands previous findings regarding body composition and cardiac health, implying that the association of lean and fat mass on long-term CVD incidence varies according to sex and prevention stage.