1. A comparison was made of the output per hectare and per cow of 3 systems of grazing management each based on a cycle of 5 days grazing followed by a 20-day rest period. The treatments were: C, strip grazing within a paddock, varying the area allotted so that a similar quantity of herbage was available each day, the pasture being frequently trimmed; treatment R, restricted grazing; the cows being allotted a fixed area of the paddock each day (67–5 m2 per cow per day) which was not trimmed or cut after grazing; and treatment RS, as treatment R but with each cow receiving cereal concentrates. Eight cows were on each treatment. Nitrogenous fertilizer was applied at the rate of 86 kg N per ha [1 kg per ha = 0.891b per acre] four weeks before the first grazing and at the same rate after each grazing; a total of 516 kg/ha was applied. The feed intake of each cow was measured on six occasions during the experiment.
2. The experiment continued for 150 days and the mean stocking rate on treatment C was 4–09 cows per ha [1 ha = 2·47 acre] and on the other two treatments 5·92 cows per ha. The number of cow grazing days per hectare were 610, 890 and 890 for treatments C, R and RS respectively and the outputs of milk were C 8450, R 12 420 and RS 13 380 kg/ha. Allowing for the concentrates given the milk outputs were C 7440, R 10 350 and RS 6660 kg/ha.
3. There were only slight differences between the treatments in output per animal.
4. The average intakes per kg W0·73 were similar in treatment C and R but the average intake of cows on treatment C was more uniform over the season. Cows on treatment RS ate consistently more digestible organic matter than those on treatment C and R but the differences became smaller as the season progressed.