1. The method of establishment markedly affected the yield of red clover in the first harvest year, the highest yield being obtained when a cover crop was used, in the absence of which red clover yield was reduced by frequent grazing during establishment. White clover made little growth in the first harvest year; the yield was best without a cover crop under frequent grazing during establishment.

2. Nitrogen applied in the seed bed did not affect the growth of red clover in the first harvest year, but depressed the white clover under infrequent defoliation without a cover crop. There was no significant effect under the other management.

3. The total yield of red and white clovers was lower, when sown with the rapidly establishing ryegrass or ryegrass-dominant general-purpose mixtures, than with cocksfoot or timothy mixtures. Weed growth, although not prominent, was far greater in the timothy and cocksfoot swards than in those based on ryegrass.

4. A close parallel existed between the growth of red clover and the dry matter and yield of crude protein. The highest figures were recorded when management during establishment had favoured the growth of the red clover.

5. Nitrogen applied in the seed bed reduced the percentage of crude protein in all mixtures in the spring and autumn grazings, and in the hay crop of the first harvest year. Management which encouraged the growth of red clover increased the crude protein percentage of herbage in the spring and autumn, but depressed it in the aftermath cut.

6. A number of interactions occurred when the relative yields of dry matter and crude protein of mixtures in the first harvest year were modified by treatments applied during establishment. There are significant indications that the method of establishment is a factor to be considered in evaluating various mixtures.