On three closely supervised stud farms, examinations for ovulation (Da
y 1) were made at intervals of 24 or 48 hours. Examination for pregnan
cy began at Day 11 to 14 and a total of 318 pregnancies were diagnosed
. Multiple ovulations occurred in 117 (36.8%) pregnant mares including
five (1.6%) triple ovulations. Sixty-two (19.5%) pregnancies were mul
tiple including four (1.3%) triplets. Forty-seven (85.5%) multiple pre
gnancies were manually reduced successfully (pregnant at Day 40). Most
twin vesicles which failed after manual reduction were small-for-age.
Seven sets of twins in which one vesicle was small-for-age were succe
ssfully allowed to undergo natural reduction, which was complete in ea
ch case by Day 21. A vesicle was not identified in seven (2.2%) pregna
nt mares on Day 13 nor in four (1.3%) mares on Day 14. When identified
subsequently, all vesicles in these mares were small-for-age and of t
hose not seen on Day 13 only three (43%) survived whereas none of the
four vesicles not identified on Day 14 survived. Embryonic losses were
not recorded from pregnancies identified on Day 11 or Day 12. It is c
oncluded that examination for pregnancy at the earliest opportunity id
entifies a high incidence of multiple pregnancy, eases the choice of e
mbryo for manual reduction and may allow prediction of early embryonic
failure of single pregnancies.