The field determination of N2 fixation by 15N‐isotopic methods provides a direct and precise measure of N2 fixation. Since 15N‐enriched fertilizers are expensive, 15N‐depleted (NH4)2SO4 was used as a more economical, readily available alternative to determine N2 fixation in common bean, Phaseolus vulgaris L. The objectives of this study were to determine whether 15N‐depleted (NH4)2SO4 provided reliable estimates of relative differences in N2 fixation among bean lines comparable to those obtained using 15N‐enriched (NH4)2SO4, and whether nine previously selected progeny lines differed from their respective recurrent parents for N2 fixation ability. The nine progeny lines and three parents were grown in two randomized complete block field experiments that were treated identically except that each experiment received only one of the two 15N fertilizers. Lines were harvested at an early reproductive stage (R3) and maturity (R9), partitioned into plant parts, and the 15N and total N yield of the tissue determined. The two 15N fertilizers produced comparable differences among lines at the R9 stage based on significant genotypic rank correlations between the two experiments, and similar fertilizer use efficiency values (% FUE). Three of the five selected progeny lines from population 24 fixed significantly more N2 than their recurrent parent ‘Sanilac’. The four progeny lines from population 21 were not significantly different from or less effective in N2 fixation than their recurrent parent ‘Porrillo Sintetico’, therefore, either previous selection based on indirect measures of N2 fixation was not effective, or the two parents did not differ genetically for N2 fixation. In all progeny and parental lines, seed N2 fixed was positively rank correlated with plant N2 fixed, and the highest N2‐ fixing lines in both populations had more fixed N2 in the seeds than did low N2‐fixing lines.