The effects on clover and grass growth of five levels of ammonium nitrogen (NH4-N) applied before sowing, at sowing and after nodule initiation have been investigated. A pot experiment in which SI84 white clover was grown in a peat soil showed that NH4-N up to 688 mg N per pot (approx. equivalent to a field rate of 120 kg ha~' N) applied before sowing and at sowing did not affect clover growth. N-fixing activity (C2H2reduction), on the other hand, was reduced progressively up to the highest level (688 mg N per pot). Application after nodule initiation increased growth relative to the zero-N treatment at all levels of application. Maximum growth and N fixation occurred at 516 mg N per pot where the DM yield was 70% higher than in the absence of added N. A field trial in which SI84 clover and S24 perennial ryegrass were surface sown on to a peat soil showed an increase in grass and clover growth in the first year in response to 120 kg ha ~' N applied at sowing. Grass growth alone was increased at 120 kg ha"' N applied 40 d before sowing. Lower rates of application before sowing and at sowing did not affect clover or grass growth. The effect of the delayed application of NH4-N on legume growth was less marked than that in the pot experiment, 90 kg ha~' N stimulating clover growth by 40% in the first year. The effect was however different from that in the pot experiment, in that, whilst 30 kg ha"' N increased N fixation relative to the zero-N treatment, plants exposed to higher levels showed a depression in