Five field experiments measured the effect of three sources of nitrogen (N) fertilizer, applied at 45 kg N/ha, on the incidence of take-all and grain yield of wheat. The N fertilizers were ammonium sulphate, ammonium chloride and sodium nitrate. Compared with the nil N treatment, ammonium nitrogen fertilizer, either as ammonium sulphate (ASdr) or ammonium chloride (ACdr) drilled with the seed, lowered the severity of take-all. Sodium nitrate topdressed (SNtd) to the soil surface reduced the severity of take-all in three of the five experiments, while ammonium sulphate topdressed (AStd) reduced the severity in four experiments. Ammonium sulphate and ammonium chloride drilled with the seed were equally effective in reducing the severity of take-all in three of the five experiments. However, ACdr was more effective than ASdr in reducing the severity of take-all in one experiment, whereas ASdr was more effective than ACdr in another experiment. In experiments 1 and 5, the reduction in take-all severity between the ASdr and ACdr treatments did not affect grain yield.