The role of alfalfa (Medicago sativa L.) in extracting NO3–N from deep soils of areas with cold, short growing seasons, such as western Canada, is not well understood. A study was established in 1990 to determine NO3–N extraction ability to 300 cm; initial soil NO3–N concentrations were high (>8 mg kg−1). Systems included continuous alfalfa; annual rotations of spring wheat (Triticum aestivum L.), field pea (Pisum sativum L.), and barley (Hordeum vulgare L.); a native‐grass system [big bluestem (Andropogon gerardi Vitman) and western wheatgrass (Agropyron smithii Rydb.)]; and continuous fallow. The annual rotation effectively lowered NO3–N to <2.3 mg kg−1 in the 30‐ to 90‐cm depth. By the 4th yr, alfalfa had reduced NO3–N concentrations to <3.8 mg kg−1 for the 30‐ to 240‐cm increment. The greatest NO3–N extraction benefits of alfalfa were realized in the 4th yr at a maximum soil depth of 270 cm. Subsoil NO3–N concentration increased in the continuous alfalfa between the 4th and 6th yr. Greater NO3–N extraction occurred with the native‐grass treatment compared with continuous alfalfa in the 0‐ to 120‐cm soil depth; however, similar extraction patterns existed below 120 cm. A system involving 4 yr of alfalfa followed by two wheat crops resulted in the lowest subsoil NO3–N concentration, even lower than the continuous alfalfa and native‐grass systems. It was concluded that subsoil NO3–N extraction with alfalfa was maximized when alfalfa was rotated with annual crops.