Legume rotations and animal manures can reduce synthetic fertilizer use but may affect emissions of greenhouse gases and ammonia in a row cropping. In 2006-2007, soil-atmosphere fluxes of N₂O, NH₃, and CO₂ were measured in maize (Zea mays L.) crops within a replicated long-term experiment comparing continuous maize to maize–alfalfa (Medicago sativa L.) rotations. In both systems, comparisons of synthetic fertilizer N and manure N were initiated in 1990. With inorganic fertilizer as the main N source, mean CO₂–C fluxes were lower from continuous maize (CC, 512 ± 132 g m^–2 growing season^–1) than from maize following alfalfa (CA, 691 ± 91 g m^–2 growing season^–1). In contrast, with manure as the main N source, mean soil CO₂–C fluxes from CC (943 ± 111 g m^–2 growing season^–1) were greater than from CA (682 ± 21 g m^–2 growing season^–1). Soil CO₂–C emissions correlated with long-term inputs of manure. Synthetically fertilized continuous maize had lower N₂O-N fluxes (0.36 ± 0.26 g m^–2 growing season^–1) than other treatments (0.55–0.58 g m^–2 growing season^–1). Nitrous oxide-N fluxes were not correlated with current N inputs or soil nitrate concentrations, suggesting that long-term treatment effects (e.g., on soil structure, labile C, or microbial communities) contribute to contemporary N₂O variation. Elevated NH₃ fluxes (>5 mg NH₃–N m^–2 h^–1) followed manure applications, but within weeks there were no significant treatment differences in NH₃ fluxes. This study shows that short-term or single-factor experiment may not capture important interactions among crop rotations and N sources affecting greenhouse gas emissions from agricultural soils.