SUMMARYThe potential benefit to be derived from seed inoculation of Phaseolus vulgaris beans with effective strains of Rhizobium phaseoli, was investigated in field experiments over three years on a site low in soil nitrogen and lacking indigenous effective strains of R. phaseoli.Inoculation with R. phaseoli (strain RCR 3644) produced significant increases in nodulation, nitrogenase activity and plant growth in all experiments. In trials in 1978 and 1979, with cv. Seafarer, inoculation, in the absence of nitrogen fertiliser doubled seed yields. In 1978, the seed yields from inoculated beans without nitrogen fertiliser (1–6 t/ha) were not significantly different from those obtained with uninoculated beans receiving the optimum nitrogen fertiliser treatment of 120 kg N/ha (1–75 t/ha). In 1979, with lower rainfall favouring more efficient utilisation of nitrogen fertiliser, inoculation gave seed yields (1–88 t/ha) equivalent to those obtained with 60 kg N/ha (1–70 t/ha) but significantly less than with 120 kg N/ha (2–88 t/ha). More precise estimates from nitrogen response curves showed that inoculation supplied the fertiliser equivalent of 105 and 70 kg N/ha in 1978 and 1979 respectively. In both years, significant benefits were also obtained by the combination of inoculation and nitrogen fertiliser.In a separate experiment in 1979, with four R. phaseoli strains inoculated onto eight bean cultivars, three were highly effective nitrogen fixers on all cultivars. Two strains (RCR 3644 and NVRS 963A) each increased mean yields, in the absence of nitrogen fertiliser, from 1–39 t/ha uninoculated to c. 2–5 t/ha inoculated whilst strain RCR 3622 was outstanding with a mean yield of 3‐0 t/ha. An analysis of the nitrogen content of seed showed that gains from nitrogen fixation were 37–57 kg N/ha/growing cycle for the combination RCR 3644 with cv. Seafarer. However, 106 kg N/ha/growing cycle was recorded for the combination RCR 3622 and cv. Aurora.