A 10-yr designed study on crossbreeding compared a 3-breed rotational crossbreeding program (ProCROSS) with Holstein (HO) cows for health treatment cost during parities 1 to 3. Crossbreds and HO cows were housed in 7 high-performance dairies. Two-breed crossbreds were Montbéliarde × HO (MH, n = 576) and Viking Red × HO (VH, n = 604) versus HO (n = 1,239) cows that calved for a first time from 2010 to 2015. Three-breed crossbreds were Montbéliarde × VH (n = 386) and Viking Red × MH (n = 401) versus their HO contemporaries (n = 871) that calved a first time from 2012 to 2016. A fixed treatment cost including veterinary fees, pharmaceutical expense, and labor cost was assigned to 14 types of health treatments. Total health cost (THC) was summed across each lactation and subsequent dry period for each cow. The 2-breed and 3-breed crossbreds were compared with HO separately with a phenotypic mixed model including the fixed effects of parity, herd-year-season of calving nested within parity, breed, crossbred group nested within breed, interaction of parity and breed, interaction of parity and crossbred group nested within breed, random sire effect nested within sire breed, and random cow nested within breed group. All fixed effects were statistically significant except for some interactions that were required to estimate solutions. The 2-breed cows had mean THC of $34, $49, and $73 during their first 3 parities, respectively; however, the HO cows had significantly more (P < 0.01) mean THC of $45, $66, and $93, respectively by parity, compared with the 2-breed cows. For the 3-breed cows, the crossbred ($33) and HO ($40) cows did not differ (P = 0.31) for mean THC during first lactation. During second ($60 vs. $81) and third ($80 vs. $136) lactations, mean THC was significantly less (P < 0.01) for 3-breed crossbreds vs. HO, respectively. The reduction of THC for 2-breed (−22% to −24%) and 3-breed (−18% to −41%) crossbred cows vs. HO cows should increase the profitability of crossbreeding.