Abstract Two factorial fertilizer trials involving four levels of nitrogen and four levels of potassium were carried out on sweet potato to gain a better understanding of the way in which fertilizer influences growth and yield. The trials were done on a Vitrandept soil at Keravat which has a wet lowland tropical climate. The general pattern of crop development and yield responses is described, using the growth analysis technique. Under conditions of continuous cropping, nitrogen (N) had a greater influence on growth and yield than did potassium (K). N fertilization up to a rate of 225 kg N ha−1 increased tuber yield, mean tuber weight, total plant dry weight, dry weight of all plant components, leaf area index, leaf area duration, number of leaves per plant, mean leaf area per leaf, crop growth rate; and, for some periods, leaf area ratio (LAR) and relative growth rate. N reduced number of tubers per plant early in the crop. K fertilization up to a rate of 375 kg K ha−1 increased tuber yield, number of tubers per plant, mean tuber weight, total plant dry weight, mean leaf area per leaf and harvest index. It reduced the percentage dry weight of tubers and LAR in Trial 2. It is suggested that N influenced yield by increasing leaf area duration which in turn increased mean tuber weight and hence tuber yield. K influenced tuber yield via an increase in the proportion of dry matter diverted to the tubers and a rise in tuber number per plant. The effect of K occurred by 7 weeks after planting and it is suggested that K fertilizer be applied early in the crop life.