The limitation of symbiotic nitrogen fixation due to P deficiency restricts the development of a sustainable agriculture, particularly in Mediterrancan and tropical soils. Common bean genotypes, APN18, BAT271, PVA846, POT51, G2633 and G12168, were grown in an aerated N‐free nutrient solution at low (72 μmol plant ‐1 week ‐1 ) and control P supplies (360 μmol plant ‐1 week ‐1 ). Nitrogenase activity was estimated by in situ measurements of acetylene reduction activity (ARA) in a flow‐through system. During the assays, maximum values of ARA (peak ARA) were reached between 20 and 30 min after exposure to C 2 H 2 , depending on P treatment and growth stage. Thereafter, a decline in C 2 H 4 evolution was observed. This decline was most pronounced in low‐P plants and there was a significant genotypic effect. ARA per plant was decreased by P deficiency, mostly because nodulation was delayed and the number and mass of nodules were reduced. The ARA decrease during pod filling was also activated by P deficiency. ARA per g dry weight nodule was increased by P deficiency in G2633 and G12168, unchanged in APN18, BAT271 and POT51 and decreased in PVA846. Except for the climbing type IV G2633, total N at harvest for all P treatments was correlated with the cumulative value of peak ARA and with peak ARA at early pod‐filling which was the highest peak ARA throughout the growth cycle of type III bushy genotypes. We conclude that if phenology and growth habit are carefully considered, peak ARA is a reliable screen of genotypes for N 2 fixation tolerance to P deficiency. Selection of lines with early nodulation under P deficiency is also advisable, and the effect of P deficiency on the nodule functioning needs to be considered.