This study investigated some aspects of breeding management in the Italian Heavy Draught Horse breed, aiming at improving its efficiency at stud farm level. A first aim was to evaluate the risk of unsuccessful reproduction in mares after an early (3 years) or normal (4 years) age at first foaling, in interaction with different stud rearing systems. A second objective was the examination of the mean time length in which young 2-year-old stallions maintain a genetic superiority on older proven stallions, identifying a ‘genetic lifespan’ in which young stallions can be safely used for reducing the cost of services. Reproductive performance at first and second foaling of 1513 mares were used. Mares had a normal first foal at 3 (n = 745) or 4 years of age (n = 768) in stud farms on the basis of stable (n = 488), feral (n = 345) or semi-feral (n = 680) rearing systems. Logistic regression analysis was performed by modeling the risk of unsuccessful reproduction in the subsequent season (i.e., results at second foaling), as affected by the interaction of age at first foaling × rearing system (six classes). Genetic lifespan of young stallions was estimated by regressing the least square means from a mixed model analysis for repeated measures of individual differences in ‘total merit’ estimated breeding values (EBVs) between young stallions (mean no. of 45/year) and the mean EBV of all proven stallions in a given year of genetic evaluation (mean no. of 483/year). Young stallions born between 1999 and 2005 were used, following each generation (i.e., birth year) from 2 to 7 subsequent yearly genetic evaluations. In comparison with the best reproductive success of second foaling at 4 years in stable systems, the greatest risk of unsuccessful reproduction was at 3 years in feral (+167%) and 3 years in semi-feral conditions (+91%). Young stallions showed a 0.50 s.d. greater EBV at the first evaluation than proven stallions, with a mean annual decrease in EBV of 0.07 s.d./year on proven stallions. Optimal breeding management could be obtained in stud farms by limiting foaling at 3 years, particularly in feral and semi-feral rearing systems, and using young stallions for 3 to 4 years to maintain a perceptible selection differential with older proven stallions and to reduce cost of services. Later, the selection differential with proven stallions become less consistent and genetic improvement could be slowed down.