Nitrous oxide (N 2O) emissions from different soils under grass were measured after treatment with cow dung and urine in field trials conducted during two separate seasons and in laboratory incubation experiments. N 2O emission rates were much higher during autumn-winter than during spring-summer, and in the case of well-drained soil were substantial for both excreta types (207 mg N 2ON kg −1 of deposited dung and 197 mg N 2ON kg −1 of urine in autumn-winter). The corresponding data for poorly-drained soil were 0.2 mg (dung) and 148 mg (urine). Emissions continued over much longer periods (∼ 60 days) from sandy and stony loams than from a silty clay loam (∼ 30 days) under both field and laboratory conditions, and were not solely dependent on soil NO 3 − or NH 4 + status but also related to other factors including soil moisture, rate of plant growth and carbon availability. Results suggest that N 2O production occurred during both nitrification and denitrification processes. Emission rates of up to ∼ 1590 μg N 2ON h −1 m −2 occurred in the field, while small rates of deposition to the soil were occasionally observed. Under laboratory conditions, similar treatments produced large emissions from loam soils having pH of 4.5–6.5 and zero emissions from a peat soil with pH of 3.8. The ratio of nitrogen released as N 2O to the amount of N excreted by the livestock varied from ∼ 0% (summer) to 0.8–2.3% (winter), consistent with loss rates observed for mineral fertilizers.