There is increasing demand for sources of energy and non-meat protein with balanced amino acid profiles worldwide. Nuts are rich in protein and essential amino acids, and have a high energy value due to their high fat content. Kernels from two wild fruits in Mozambique, Adansonia digitata and Sclerocarya birrea, were selected for this study to determine their amino acid and fatty acid composition, as well as the monomeric composition of their dietary fibre. The fat content in Adansonia digitata kernels was around 32% and in Sclerocarya birrea kernels 49%. Both kernels were rich in unsaturated fatty acids. The dominating fatty acids in Adansonia digitata kernels were palmitic, linoleic and oleic acid; varying from 25.7% to 34.9% of the total fatty acids content. In Sclerocarya birrea kernels the main fatty acid was oleic acid, 72.4%. The Adansonia digitata kernels contained the two essential fatty acids linoleic (around 30%) and linolenic acid (2%). Sclerocarya birrea kernels contained linoleic acid (around 7%). The protein content in Adansonia digitata kernels was 35% and in Sclerocarya birrea kernels 29%. Glutamic acid was the most abundant amino acid, comprising more than 20% of the protein in both kernels. The contents of essential amino acids in the kernels were compared with the requirements stated by the WHO, and the findings indicate that Adansonia digitata and Sclerocarya birrea kernels can provide good, cheap sources of protein, especially when combined with foods with high lysine content. Regarding the dietary fibre fractions, the main constituent in the insoluble fraction of Adansonia digitata kernels was glucose, while in the soluble fraction arabinose was the dominating component. In kernels of Sclerocarya birrea, uronic acids constituted more than 90% of both dietary fibre fractions. The results of this study suggest that intake of these kernels can help providing a great part of the fatty acids and amino acids required in the daily diet, especially for people living in rural areas of Mozambique. The data could be used for intake estimates, and to encourage increased consumption and utilization of these kernels.