Two methods of placement of 32P in soil were used to investigate the root activity in perennial ryegrass and mixed ryegrass-white clover swards. The placement of isotope by either horizontal or vertical insertion in the mixed sward had no significant effect on the distribution of root activity with depth, the total dry-matter yield of herbage or its distribution between ryegrass and clover. At least 90% of the root activity to a depth of 35 cm for both grass and clover occurred within the upper 12.5cm of the soil profile. Irrigation resulted in a more even distribution of root activity with depth and increased the lateral exploitation of soil P. In the irrigated plot, the root activity of both grass and clover above the line of isotope placement accounted for 40-50% of the total activity at depths to 18.8cm in contrast to 70-100% in the absence of irrigation. The method of defoliation (cutting versus grazing) had no significant effect on the vertical or lateral distribution of root activity. In unirrigated mixed