Long‐term management of croplands influences the fluxes and sources of nitrous oxide (N2O). We examined this premise in a greenhouse study by using soils collected from a 38‐yr‐old field experiment. The sampled treatments were continuous barley (Hordeum vulgare L.; CB), continuous fescue (Festuca rubra L., F. arundinacea Schreb; CF), and two phases of an 8‐yr rotation: faba bean (Vicia faba L.; FB) and alfalfa (Medicago sativa L.)–bromegrass (Bromus inermis Leyss) hay. Barley was grown as a test crop in the greenhouse in each soil. The ranking of N2O emissions was hay > FB > CB > CF (P < .001). We quantified the 15N‐site preference to assess the N2O‐producing processes. Denitrification was the predominant source, contributing 77.4% of the N2O production. We also evaluated nitrogen (N) additions: urea alone or urea with a nitrification inhibitor (nitrapyrin or DMPSA). Compared with urea alone, nitrapyrin and DMPSA reduced N2O emissions by 16 and 25%, respectively. We used urea labeled with 15N to trace N to N2O emissions, aboveground plant N uptake, and N retention by soils. Total 15N‐recovery (N2O + plant + soil) was highest under FB (86%) and lowest under CB (29%). We further separated the N2O derived from urea versus N2O from soil organic matter (SOM). The inhibitor DMPSA reduced the N2O derived specifically from added urea‐N by more than half (P < .001). With the addition of urea, N2O production from mineralization of SOM‐N accelerated over the control (without urea), termed the priming effect. This priming of SOM‐N contributed with 13% of the total N2O production when averaged across the four management legacies. The CB soil had the highest proportion of priming‐derived N2O (24%). Management legacies clearly differed in soil carbon and N, which governed N2O production from denitrification and SOM priming.