The objective of this study was to develop a model for the simulation of milk production, milk composition (including milk fat UFA concentration), the segregation of dairy cows and the economic performance of dairy farms under New Zealand farming conditions. The model developed was used to investigate the effect on farm production and profit of the phenotypic segregation of cows for the production of milk fat with high UFA concentration. The model used the Cholesky decomposition algorithm of (co)variance matrices to simulate the performance of Holstein-Friesian cows for milk yield (MY), fat percentage (F%), protein percentage (P%), fat UFA concentration and live weight (LW). The mean performance of cows and farms simulated by the model were very close to national average statistics for New Zealand dairy farms. The model was used to simulate: (1) a population of 1,820,000 cows in 5600 farms (AVE farms), and (2) the establishment of a farm (120 ha) for the production of milk high in UFA through the segregation into a herd of the top 325 (2.71 cows/ha) or 353 (2.94 cows/ha) cows for fat UFA concentration (UFA(2.71) farm and UFA(2.94); farm, respectively). The simulations were repeated 1000 times and 95% confidence intervals were estimated by bootstrapping methodology. On average, the UFA(2.71) and UFA(2.94) farms produced milk with 23.6% more UFA than AVE farms. However, cows on the UFA(2.71) and UFA(2.94) farms had significantly lower yields of fat (both -48 kg, P < 0.05), protein (-24 and -23 kg, respectively, P < 0.05) and milksolids (-73 and-72 kg, respectively, P < 0.05) than cows on AVE farms. Under a milk payment system that pays for yields of fat ($3.80/kg) and protein ($9.67/kg), and penalises milk volume (-$0.03/1), the UFA(2.71) and UFA(2.94) farms had significantly lower operating profit (-$872/ha and $946/ha, respectively, P < 0.05) than AVE farms. These results indicate that farm profit would be adversely affected unless there is a premium for fat UFA concentration. (C) 2016 Elsevier B.V. All rights reserved.