Abstract: The objectives of the reported research were to compare, using gross margin, dominance and marginal analysis, summer legume-wheat (Triticum aestivum) and fodder maize (Zea mays)-wheat systems in northern Pothwar, Pakistan. Legume species were soybean, mungbean and black gram. Data on legume shoot and grain yields of dry matter (DM) and nitrogen (N) and N2 fixation, using natural 15N abundance, were combined with grain yields of subsequent wheat crops for inclusion in the economic analyses. Mean grain yields for the three legumes were 2.7 (soybean), 0.4 (mungbean) and 1.1 t ha-1 (black gram). Maize not fertilized with N produced 7.2-8.9 t ha-1, compared with 5.8-13.2 t ha-1 when fertilized. Estimates of the proportion of legume N derived from the atmosphere (%Ndfa) were 31-64 % for soybean, 37-44 % for mungbean and 42-48% for black gram. Both legume rotation and fertilizer N significantly improved the yields of following wheat crop. In the absence of fertilizer N, increases were 15-70% for DM and 14-67% for grain yield, relative to maize. When the wheat was fertilized with N, the legume benefits disappeared. Responses of the wheat to fertilizer N, irrespective of prior crop, were large, ranging from 75 to 276%. Gross margin analysis indicated the black gram/N-fertilized wheat and soybean/N-fertilized wheat sequences to be the most profitable, with the large variation in gross margins (Pakistani Rs. 2861 to Rs. 115500) reflecting the wide range of yields and grain prices. Dominance analysis confirmed the economic benefits of black gram-based sequences, particularly at low to intermediate levels of input. These results highlight the need for thorough economic, as well as biological, analysis of cropping options.