A group of 143 dogs and cats with pruritic skin disease was tested by intradermal injection of a dilute whole cat flea extract. The test results were correlated with the clinical diagnosis (Table I, III), with previous or present findings of fleas on the animals (Table II, III), with the presence or absence of blood eosinophilia (Table IV), and, in animals with clinical flea allergy, with the histological findings in a skin biopsy excised from a clinically affected skin site (Fig. 1). The predictive value of a positive and a negative outcome of the test was assessed graphically by means of calculations described in the Appendix (Figs. 2, 3). It is concluded that the test is sufficiently specific to justify its use in animals where flea allergy is suspected but cannot be proved by routine clinical examination. The histological findings and the nature of the skin reactions suggest that a positive outcome of the test is indicative of an immediate, type I hypersensitivity towards components of the oral secretion of fleas and that this hypersensitivity is mediated by reaginic antibody which to some extent is produced by plasma cells located in skin areas where flea saliva is deposited.