Translocation and assimilation of N during grain filling are involved in determining grain protein content in wheat (Trificum aestivum L.). Considerable controversy on the physiological basis for high grain protein concentration exists. These aspects of N movement within the plant were studied in field experiments with the spring wheat cultivars Anza and Cajeme 71 and 96 F5 lines from the cross of these cultivars. Cajeme 71 had higher grain protein concentration, higher translocation, translocation efficiency, and a higher proportion of N in the grain than Anza, but no relationship between any of these three parameters and grain protein concentnation was found in the F5 lines. The ratio of N harvest index to grain harvest index, which gives a rough estimate of the relative proportioning of N and carbohydrate in the grain, was positively related to grain protein concentration (r=0.39 to 0.45, P < 0.01), emphasizing the need to consider both N and carbohydrate partitioning while studying the genetic basis of high grain protein concentration. A relatively simple genetic basis was postulated for the observed variation. A two‐gene additive model fit the F5 frequency distribution observed in both low and high N experiments. The F5 lines that assimilated more N after anthesis than required for their yield level, identitied by regression of N assimilation after anthesis on grain yield (Le., high positive residuals), and that had high N translocation were found to be high in grain protein concentration. Excess assimilated N, the sum of translocation and these residuals (positive or negative), was closely related to grain protein concentration (r=0.8 to 0.9, P < 0.01). Further, these data for excess N conform closely to theoretical calculations of the additional N requirement for a 1% increase in grain protein concentration.