Common laboratory mouse strains have very low complement levels relative to humans, rats, guinea pigs, rabbits and other mammals, which limits the value of the mouse as an experimental model. We therefore tested serum complement levels of 43 mouse strains and 11 rat strains, for the purpose of selecting a convenient laboratory animal having high complement levels. Total complement activity was determined with both erythrocytes and human tumor cells as targets. Eight mouse strains were identified that have complement levels comparable to those of other mammals. These mouse sera lyse tumor cell targets as well as sera from humans, rats or guinea pigs, although they are somewhat less active than rabbit sera. They are relatively inefficient in lysing erythrocyte targets, yet are as active as rabbit serum in this assay. Target cell lysis was demonstrated to be via the classical pathway of complement activation. Of the eight 'high complement' mouse strains, four were recently derived from wild mice, and one, SF/CamEi, was derived from wild mice in 1951. The three other strains, BUB/BnJ, DA/HuSn and RIIIS/J, were developed more than 40 years ago, but apparently were not tested previously for complement activity. Using the BUB mouse as a representative of the 'high complement' mice, we assayed levels of the nine complement components, in an attempt to identify the cause of high complement activity. No difference in levels of C1, C2, C4, C8 or C9 was detected between BUB and BDF1 mice. C2 activity was very low in both strains. C3, C5, C6 and C7 activities were higher in BUB mice than in BDF1 mice, indicating that variation in these complement components is responsible for the difference in total complement activity. The genes determining the 'high complement' phenotype appeared to be semi-dominant in F1 hybrids. The 'high-complement' mouse strains, and recombinant strains derived from them, will be useful in a wide range of biomedical research.