Ninety-four cows were randomly allocated to 1 of 5 stocking rates (2.2, 2.7, 3.1, 3.7, and 4.3 cows/ha) in a completely randomized design for 3 years. Herds were seasonal calving, with only minor differences in grazing management to optimize the profitability of each stocking rate (SR). Pasture production and quality data, milk and milk component data, and reproduction data were collected, averaged for SR treatment, and linear and quadratic contrasts on SR were evaluated. In addition, the Wilmink exponential model (y(t) = a + b x e((-0.05t) )+ c x t) was fitted to milk yield within lactation, and the parameters were averaged by SR treatment and analyzed as above. The median variation explained by the function for individual lactations was 84%. The amount of pasture grown tended to increase, and the quality of the pasture on offer increased linearly with increasing SR, reducing some of the negative impact of SR on the availability of pasture per cow. Milk production per cow declined linearly with increasing SR, although there was a tendency for most production variables to decline quadratically, with the negative effect of SR declining with increasing SR. The effect on milk production per cow was primarily because of a lower peak milk yield and a greater post-peak decline (less persistent milk profile), although a decline in lactation length with increasing SR was responsible for 24% of the effect of SR on milk yield. Milk production per hectare increased linearly with increasing SR, and there was only a small difference (approximately 3%/cow per ha) in the efficiency of converting feed dry matter into milk energy. Stocking rate did not affect reproductive success. The data are consistent with the need for a more robust measure of SR than cows per hectare because farms will differ in the genetic merit of their cows and in the potential to produce pasture. We introduce the concept of a comparative SR, whereby the carrying capacity of the farm is defined by the BW of the cows, the potential of the land to produce pasture, and the amount of supplement purchased (kg of BW/t of feed dry matter). The adoption of such a measure would facilitate the extrapolation and transfer of research findings among systems.