Two experiments with dairy cows and one with suckler cows and their calves were conducted to examine the use of secretion rate measurements for estimating total milk production. In the first experiment both 4- and 6- h intervals between measurements gave similar estimates of total 7-d milk yield. The second experiment compared estimated and measured milk composition as well as yield. Milk and solids-not-fat yields were underestimated with dairy cows as a result of an extended milking interval before measurement. However, fat yield was overestimated, indicating that all residual milk was not removed at the first oxytocinaided milking. It was concluded that for the beef cow, previous interval effects would be eliminated by the frequency of calf suckling, but that residual milk effects might cause a 3–6% and a 16% overestimation of milk and fat yields respectively.

In the third experiment, the milk yield of suckler cows was estimated from measurements of secretion rate and from changes in calf weight; good agreement was obtained provided there were at least 3 consecutive controlled sucklings.