Is there a way to provide crops with enough nitrogen (N) to meet production objectives while simultaneously limiting the environmental effects of N fertilisers? At sites in western France, we studied the multiyear impacts of using Italian ryegrass as a catch crop in a wheat-maize rotational system. A 14-year and a 12-year experiment were conducted; in the latter, the effects of a liquid-manure treatment were examined. We analysed changes in maize and wheat yields, optimal levels of added N, nitrogen mineralisation dynamics, and soil stores of organic C and N. In some cases, the catch crop had a positive long-term effect; the plant's relative impact depended on its level of growth and, therefore, the amount of N it absorbed and subsequently returned to the soil. Making N available to succeeding crops had variable annual effects; consequently, it may not be helpful in reducing the nitrogen dependence of crop systems over the short term.