1. Laboratory rabbits were infected with seventeen different preparations of myxoma virus (this probably represents thirteen strains of which four could be classified as fully virulent and the rest as attenuated, two being markedly so).
2. One hundred and twenty-four wild rabbit carcasses were obtained from a rabbit infestation in the throes of an outbreak of myxomatosis.
3. All the carcasses were examined for the presence or absence of myxoma antigens and antibodies by a modified Ouchterlony gel-diffusion precipition test.
4. As a result of (3), we have concluded that the gel test can provide confirmation of the presence or absence of myxomatosis. It is particularly useful in cases of doubt about the clinical diagnosis. The test can also give a reasonable indication of the number of recovered rabbits in a population and, more important still, it can give an indication of the recovery rate during an outbreak of disease. However, attempts to differentiate between fully virulent and attenuated strains of virus were unsuccessful.