Using an experimental method of half-udder comparisons the effects on the amount of residual milk of varying milking intervals, the duration of milking, and the delay between the stimulation of milk ejection and the start of milking were measured. Yield of residual milk was not related to the length of milking intervals within the range of 4–16 h but the butterfat content was directly related to the length of the interval. Neither the yield nor the butterfat content of residual milk was affected by a change in the duration of milking of 1 min. A delay between the stimulation of ejection and the start of milking of 6–8 min resulted in increasing the yield of residual milk by about 1 lb.

It is suggested that these results indicate that milk ejection is sufficiently persistent that residual milk is unlikely to be affected by delays between stimulation of milking and the start of milking until they exceed 4–6 min.