We sought to evaluate the effectiveness of a policy of early elective hospitalization on the outcomes of 522 consecutive twin gestations delivered at our institution between 1983-1987. During the first 2 years (1983-1985), 237 twin pregnancies were delivered with a policy of elective hospitalization when twin pregnancy was diagnosed between 24-32 weeks' gestation. When possible, elective hospitalization started at 24 weeks' gestation. Electively admitted women remained hospitalized until 34 weeks' gestation, at which time they were discharged unless complications developed requiring continued hospitalization. During 1985-1987, 285 women with twin gestations were intentionally managed as outpatients unless intercurrent complications required hospitalization. A total of 211 twin pregnancies was excluded from analysis because the women did not present for prenatal care (19%) or were undiagnosed until delivery (22%). Of the remaining 311 pregnancies available for study, 134 were managed when the elective admission policy prevailed and 177 when this policy was not in effect. Although the elective admission policy did result in a small reduction in the incidence of low birth weight among the 58 pregnancies hospitalized electively (mean [+/- SEM] gestational age at elective hospitalization 27.7 +/- 0.3 weeks) compared with outpatient management, this policy did not result in an improvement in prematurity (32 versus 36%; P greater than .05) or perinatal morbidity as reflected by requirement for neonatal intensive care (12 versus 11%; P greater than .05) and mechanical ventilation (8 versus 9%; P greater than .05). Moreover, perinatal mortality was actually higher in the electively hospitalized pregnancies (8 versus 2%; P = .01).(ABSTRACT TRUNCATED AT 250 WORDS)