Biofortified orange-fleshed sweetpotato (OFSP) is being promoted to tackle vitamin A deficiency, a serious public health problem affecting children and pregnant/lactating women in sub-Saharan Africa. The aim of the study was to quantify and understand the factors influencing carotenoid losses in dried OFSP. Losses were determined in chips after drying and storage. A preliminary study demonstrated that carotenoid levels were not significantly different following either solar or sun drying. Carotenoid loss after drying was generally correlated with high initial moisture content and high carotenoid content in fresh sweetpotato roots. Losses of pro-vitamin A were less than 35% in all cases. Flour made from OFSP could therefore be a significant source of provitamin A. In contrast, storage of chips at room temperature in Uganda and Mozambique for four months resulted in high losses of pro-vitamin A (ca. 70-80% loss from the initial dried product). Low-cost pre-treatments, such as blanching, antioxidants and salting, did not reduce carotenoid losses during storage. Enzymatic catabolism of [bêta]-carotene in dried OFSP was considered unlikely because of low peroxidase activities at low water activities and the loss of peroxidase activity during storage. To understand further the factors causing the losses, dried sweet potato chips were stored under controlled conditions of temperature (10; 20; 30; or 40°C), water activity (0.13; 0.30; 0.51; 0.76) or oxygen (0 [under nitrogen]; 2.5; 10 or 21% [air]). Oxygen was the main cause of degradation followed by temperature. An Arrhenius kinetic model was used to show that carotenoid breakdown followed first order kinetics with an activation energy of 68.3kJ.mol-1 that was in accordance with the literature. Experimental observations fitted well with data predicted by the kinetic model. The formation of the volatile compounds, [bêta]-ionone; 5,6-epoxy-[bêta]-ionone; dihydroactinidiolide; [bêta]-cyclocitral that were clearly related to