The Karoo is an arid inland plateau in the central and northeastern Cape Province of South Africa and is characterised by sparse, stunted vegetation. The vegetation is rich in species, and over large areas species of Pentzia, which are drought-resistant shrubs, are extremely abundant. These plants are food for stock (mainly sheep), and because the larvae of the Karoo caterpillar periodically occur in sufficient numbers to defoliate the plants, they assume pest status. Fully-fed Loxostege frustalis larvae construct cases in the soil beneath their food plants and they overwinter in these cases. A census of the numbers of larval cases accumulated in the soil, conducted from 1975 to 1980, showed that an expanded distribution of the pentzias is responsible for the periodic larval outbreaks. Further, alternate food plants are an important food supply for L. frustalis larvae when they disperse under crowded conditions The census data for L. frustalis were analysed by constructing partial life tables for the life-history period from case construction to moth emergence. The mortality of 'encased larvae' is useful for assessing the impact of the known L. frustalis natural enemies. The most important natural enemies were the braconids Chelonus curvimaculatus Cameron, Macrocentrus maraisi Nixon and the fungus Beauveria bassiana (Balsamo) Vuillemin. The collective responses of all the natural enemies to the density of L. frustalis encased larvae was direct, but undercompensating, so that areas with more pentzias produced more L. frustais. Strategies for immediate measures to alleviate the Karoo caterpillar problem, and for future research, are discussed. It is concluded that reduction of pentzia populations to acceptable levels and/or supplementing pastures with non-host plants of L. frustalis offer the only practical solutions to the Karoo caterpillar problem. Biological control of L. frustalis is dismissed as an option for reducing the pest status of the Karoo caterpillar