Three manures cross-labelled with N-15 in the faeces, urine and straw fractions were prepared after feeding a sheep first with unlabelled hay and then with N-15-labelled hay. Half of each manure portion was stored aerobically (composted) with heat evolution and the other half anaerobically by preventing oxygen supply. The ratio between faeces-N:urine-N:straw-N was 1.0:1.5:0.2 immediately after mixing. All manures were stored for 86 days. The composted manure lost 46%, and 53% of its initial N and C content, respectively. Losses from the anaerobically stored manure were not totally prevented but were significantly less at 18% (N) and 24%, (C). Urine-N accounted for the highest proportion of total N losses from both the aerobically and anaerobically stored manures. However, contributions of N to gaseous losses from the faeces and straw fractions became more important as the storage progressed. Thus urine-N accounted for 79% of the total losses after the first seven days of composting but only for 64% at the end of the storage period. In the anaerobically stored manure, urine-N accounted for 94% of the total N losses after 28 days and for 68% at day 86. Because of the high urine-N loss in the composted manure, the proportion of faeces-N in total N was similar to that of urine-N after storage. The anaerobically stored manure still contained more urine-N than faeces-N. II was estimated that 46% of faeces-N was mineralised during the composting of the manure but a great proportion of the mineralised N was lost. Less faeces-N (33%) was mineralised during the anaerobic storage and the relative losses of the mineralised N were lower. The differently stored manures ended up having the same C:N ratio and total N concentrations but the anaerobically stored manure had a significantly greater proportion of inorganic N compared with the composted manure. Differences in the forms of N may influence the fertiliser value of the manures after field application.