Subcutaneous injection of the larvae is the almost universally adopted means of initiating experimental infections of skin-invading roundworms but, so far, the possibility that this procedure introduses artefacts of one kind or another has not been critically studied. Experiments described in this paper were used to compare the effect of (a) injection and (b) skin application, of a small, precisely counted (‘exact’) dose of larvae. Results with two strains of s. ratti shoewed that the same proportion the dose developed to adults in the intestines of rats irrespective of the method. With the same exact dose technique it has been shown that milk-borne infection of the pups of lactating rats is not an artefact produced by injectcion. Large doses (mean 4000) of larvae of the homogonic strain of s. ratti carrying a radioactive label of 75 Se were tracked in their migration to the mammary gland following injection or skin application at two different sites on the right-hand side of nursing mother rats. The broad conclusion of earlier work in this laboratory using injection. that larvae move by a local route and not a systemic one, was supported by the results. The detailed distribution of the label and of unlabelled worms of the heterogonic strain in families was, however, different for the methods. indicating that subtle variations in pathway can be brought about by the use of injection. If migration involves the Lymphatic system, then the interpretation of immunological experiments in terms of Lymphatic anatomy must take account of such procedural effects. The extent to which these results contribute to theories of migration in Strongyloides ratti is discussed.