The intranasal infection of mice with cowpox virus (CPXV) has been evaluated as a model for smallpox infection in man. Administration of a lethal dose of CPXV allowed time for development of T-cell responses but antibodies could not be detected before death occurred. In contrast, infection with a sublethal dose was associated with an early T-cell response followed by neutralising antibodies which correlated with virus clearance. Comparison of two first generation smallpox vaccines revealed no significant differences in terms of immunogenicity, protection and post-challenge virus clearance. These studies show that the CPXV/mouse model is valuable for the initial assessment of smallpox vaccines.