The growth and N content of rhizobia-inoculated white clover growing in soils fumigated with methyl bromide and mixed with fertiliser were measured. Treatments comprised nematode-free controls, and inoculation with Heterodera trijolii cysts, for Kokotau silt loam, and with H. trijolii cysts, Meloidogyne hapla larvae, and with both nematodes together, for Egmont brown loam; two P levels were added to each soil. N mineralisation in soil similarly treated, but without growing plants, was also determined. In Egmont soil, but not in Kokotau, herbage dry mater yields and plant N were sometimes greater at the higher level of added P. Level of added P had no effect in either soil, at the end of the trial, on fresh weight of roots, number of rhizobia nodules, or numbers of H. triiolii, when present; numbers of M. hapla in Egmont soil were greater at the higher P level. All nematode treatments eventually reduced dry matter yields, plant N%, and plant N yields: in Egmont soil H. triiolii had more effect than M. hapla, and their effects when inoculated toegther were no greater than those of H. triiolii alone. At the final harvest, root weight was significantly lower in all nematode treatments, and numbers of rhizobia nodules were generally less; within nematode-infested pots numbers of nodules and nematodes were positively, and usually significantly, correlated. Fumigated soils incubated without plants produced appreciable amounts of mineral-N, mainly ammonium-N, initially more than produced by non-fumigated soil; appreciable nitrate-N was roduced only in Egmont soil inoculated with H. trijolii and the associated soil organic fragments. Amounts of N that could have been fixed by the clovers were estimated by using the soil N mineralisation data and the plant N data: they averaged 0.424 g and 0.294 g N per pot for control plants, and 0.247 g and 0.046 g N per pot for nematode-infested plants, in Kokotau and Egmont soils respectively. These nematodes can adversely affect the N-fixing capacity of white clover, and their effect should be assessed in the field, but the results do not clearly indicate the mechanism (s) responsible.401