Two units, within a mouse-breeding colony, with different breeding systems, have been compared from the point of view of general healthiness of the breeding stock. In one unit the monogamous-pair system, involving post-partum oestrus mating, was employed and in the other the harem system, with isolation of the pregnant females for each litter-down until after the post-partum oestrus. The two units were housed separately and had different staffs. The monogamous-pair unit was accommodated in four small rooms not directly connected; the harem unit in two larger connecting rooms. The methods of cageing, bedding, feeding, watering, care, culling and subsequent procedures were identical for both units and the same system of recording individual detailed information relating to health and breeding was followed for the two units. The size of the two populations studied was 1198 and 1024, respectively, the larger number belonging to the harem unit in which the mice were retained for a significantly longer period. These mice were moved more frequently and handled much more often.

Apart from a small outbreak of Tyzzer's disease (in the harem-bred unit) and sporadic cases in both units spread over the whole period of the survey—some 3–4 years—the general health of the two units was good. There were no cases of salmonella infection, mousepox, pasteurellosis or other serious infections. Throughout the period both units maintained a high standard of productivity and there was no evidence that the mating of the monogamous pairs at the post-partum oestrus led to any weakness or ill health. As would be expected, however, the losses from death and culling were greater among the females in both units.

The health record of the monogamous-pair bred unit was consistently higher than that of the harem-bred unit. There were more cases of Tyzzer's disease in the harem-bred mice and more evidence generally of ill health in this unit. Infertility was three times greater with the harem-bred mice as compared with those in the monogamous-pair unit.

Details of the two systems are discussed and although the records maintained for the two may not be strictly comparable it is concluded that, under the conditions obtaining, the monogamous-pair breeding system yielded a healthier colony of mice than the harem breeding system.

I should like to thank Dr J. S. Patterson and Mr R. Cook for placing many of their records at my disposal, Mr S. Peto and his staff for carrying out much of the tedious compilation of data, and Dr D. W. Henderson, F.R.S., for affording me facilities for this work. To these I am especially grateful for their interest, advice and assistance.