The final test of a smallpox vaccine is its capacity to prevent the disease from developing in inoculated individuals. This capacity, however, cannot be measured directly, so that other methods of assessing the efficacy of vaccine have had to be developed. A laboratory method-pock counting on the chorio-allantoic membrane of chick embryos-has recently been shown to provide a reasonably reliable estimate of the number of infective units in a given vaccine. In this paper, the author compares this pock-counting method with another method-titration by intravenous injection of chick embryos. He concludes that, although the reproducibility of titrations by intravenous injection compares very favourably with that obtained by chorio-allantoic inoculation, the former method would not be advantageous for the assay of vaccines, since it is very time-consuming and since differences in virulence might obscure comparisons between the efficacy of vaccines.