The benefits to establishment and growth of white clover cvs Aberystwyth S.184 and Grasslands Huia of inoculation with three strains of Rhizobium trifolii, using the peat or liquid inoculum techniques, were investigated during 1975-8 on improved hill soils ranging from brown earth through dry and wet peaty podzol to deep peat. Inoculation induced positive response in either number of seedlings, plant cover or dry-matter production in 18 out of 139 comparisons, had no effect in 118 and produced a negative response in three. Most of the positive responses to inoculation were at sites with wet peaty podzol or deep peat soils but of the five sites where increase in clover d.M. production was found in the first harvest year one was a brown earth. The positive agronomic responses occurred only when the proportion of plants with nodules was high and where a substantial proportion (> 50 %) of the latter contained introduced Rhizobium strains at least in the year of sowing. The three negative responses were in numbers of seedlings on one brown earth and two dry peaty podzol soils and with the Huia cultivar only. Despite lack of statistical significance at individual sites the dominant overall trend was for inoculation to enhance seedling establishment and the early growth of white clover in all soil types. On one brown earth and one dry peaty podzol soil there was some evidence that spraying the Rhizobium on to emerging white clover seedlings was more beneficial, at least in microbiological terms, than the customary peat inoculum procedure. The incorporation of even a small amount of nitrogen (30 kg/ha) into the seed bed at the time of sowing adversely affected germination, establishment and growth of white clover in some soils. Sometimes the effects of this nitrogen persisted into the first harvest year.