An evaluation programme is presented for determining activity of unknown compounds on Ascaris suum larvae migrating into pulmonary tissues of animals. Data on gross lung lesions due to migrating ascarid larvae were obtained in a series of 44 screening tests consisting of 223 group observations from 1115 mice orally infected with A. suum. Lesion scores were graded in ascending order from 0 to 5 depending upon number and extent of pulmonary haemorrhages. In general, there was no significant difference in lesion scores between untreated mice and mice treated with unknown compounds in their ration. The programme is based on an initial 95% rejection of compounds with a 5% ability to retest compounds in groups of mice.The grand mean score for 44 tests was 2·88. Ninety per cent confidence limits for means among tests established scores at 2·21 and 3·55, and among groups of mice at 1·76 and 4·00, When the test mean was less than 2·2, the ascarid ova used for infection were considered non-virulent. Compounds selected as active were in groups of mice having a mean lesion score of 1·76 or less. Banminth, Styquin and Tetramisole showed larvicidal activity. A reasonable compromise exists in this system wherein false leads are tolerated to avoid screening out effective materials. The host-parasite relationship is proposed as a valuable tool for anthelmintic evaluation.The authors wish to thank Mr W. Henry and Mr M. Jackley of this laboratory.