OBJECTIVE: To compare the effect of hCG therapy with placebo or bed rest in women with threatened abortion. DESIGN: Random assignment, double-blind. SETTING: Community hospitals. PATIENTS AND INTERVENTIONS: Sixty-one women with viable pregnancies under 8 weeks' gestation on admission to hospital were randomly allocated to receive injections of hCG or placebo, or were advised to rest in bed. MAIN OUTCOME MEASURES: Abortion vs. continuation of pregnancy at 16 weeks' gestation. RESULTS: Thirty-one aborted: 6/20 on hCG, 10/21 on placebo, 15/20 on bed rest. hCG vs. bed rest, P abortion, P < .01; continued with hCG vs. aborted on placebo, P < .001; continued with hCG vs. aborted with bed rest, P