Investigations on the possible biological control of a particular insect pest usually have as their object the discovery of parasitic or predacious species which, when liberated in a different area without their own natural enemies, will in time bring about an economic reduction in the numbers of the pest concerned. Hence it is desirable to have some standard of comparison between various parasites attacking a pest of their efficiency as controlling agents. The standard normally used is that of the percentage of hosts in a certain stage found to be attacked by a given parasite species. Samples of host material are taken in the field and either dissected or reared in the laboratory. In either case the parasite species present are noted, the percentage of the hosts parasitised by each species worked out (together with the figures for hyperparasitism, if any) and a comparison made between them as to their relative value as controlling agents. Often such estimations of the value of a parasite species are based on a few small localised collections of field material; but apart from any criticism arising from a statistical consideration of the methods of sampling there are certain difficulties of a biological nature inherent in the estimation of the degree of control exerted by individual parasite species, Some of these Will be outlined below.