SUMMARYIn five experiments made at Rothamsted from 1966 to 1970 ‘Nitro-Chalk’ (ammonium nitrate–calcium carbonate mixture, 21% N) was broadcast for grass cut for silage, either in a single dose in spring or divided equally for three cuts. The ‘Nitro-Chalk’ was used to evaluate anhydrous ammonia (82% N) in 1966, anhydrous ammonia and aqueous ammonia (approximately 26% N) in 1967, 1968 and 1969, and anhydrous ammonia, aqueous ammonia and aqueous urea (18% N) in 1970. All these fertilizers were applied to give 125, 250, 375 and 500 kg N/ha, except in 1968 when 250, 375, 500 and 625 kg N/ha were given.Anhydrous ammonia gave smaller yields (of dry matter) than the other N fertilizers except in 1968, a wet year, when it was at least as good as ‘Nitro-Chalk’, but slightly less good than aqueous ammonia. Yields were larger with autumn- than with springinjected aqueous ammonia and larger with either, than with equivalent single doses of ‘Nitro-Chalk’. Yields with aqueous ammonia were also larger than with ‘Nitro-Chalk’ divided equally for three cuts when more than 375 kg N/ha was tested, but smaller with less. Aqueous urea was as good as a single dose of ‘Nitro-Chalk’ but less good than divided ‘Nitro-Chalk’ in 1970, the only year it was tested. ‘Nitro-Chalk’ divided equally for three cuts gave larger yields than when a single dose of up to 375 kg N/ha was given, but with more N than this a single dose was better.Apparent recovery of N was least from anhydrous ammonia and differed little between autumn and spring injection in 1968, the only year when a valid comparison was possible. Apparent recoveries of N from autumn and from spring-injected aqueous ammonia and from ‘Nitro-Chalk’ broadcast in a single dose differed little, nor did the proportion of the N recovered at each cut. Recovery from ‘Nitro-Chalk’ broadcast in three equal amounts was more uniform, but little larger.Judged by yield, no more than 375 kg N/ha was justified; grass given this amount of N either as aqueous ammonia or as ‘Nitro-Chalk’ removed approximately 29 kg of N, 3 kg of P (7 kg P2O6), 26 kg of K (31 kg K2O) and 2 kg of Mg (3 kg MgO) per tonne of dry matter produced.