Information on amounts of N2 fixed by legumes in subarctic regions is lacking. We determined the N2‐fixation potential of seven legume species under field conditions in subarctic Alaska. Seasonal N2 fixation was estimated for alfalfa (Medicago sativa L.), red clover (Trifolium pratense L.), yellow sweetclover (Melilotus officinalis Lam.), fababean (Vicia faba L.), lentil (Lens culinaris Medik.), pea (Pisum sativum L.), and white lupin (Lupinus alba L.) on a neutral (pH 7.2) and an acid (pH 5.4) soil during two field seasons. A lime variable was included at the acid soil location. Dinitrogen fixation was estimated by the 15N isotope dilution method and by the total‐plant N difference method. Three non–N2‐fixing crops and uninoculated analogs at one location were used as reference crops. Agreement among reference crops and between methods of estimating N2 fixation was adequate for most applications, but differences of >50 kg ha−1 were sometimes observed. Maximum N2 fixation estimates by the isotope dilution method were 58 kg ha−1 for alfalfa, 121 for red clover, 109 for sweetclover, 204 for fababean, 72 for lentil, 121 for pea, and 162 for white lupin. Liming usually did not significantly affect N2 fixation. Nitrogenase activity, as measured by the acetylene reduction assay, peaked at or soon after flowering in some legumes; in others, it either plateaued or continued to increase after flowering until the end of the growing season.