The sexual transmission of hepatitis C virus (HCV) has long been debated. The prevalence of infected at-risk partners varies from 0% to 30%. In a prospective study, the risk of infection was quantified in steady heterosexual partners and the prophylactic effect of normal human polyvalent immune serum globulin (ISG) was evaluated. A total of 899 at-risk partners of HCV-infected patients were enrolled in a single-blind randomized controlled trial and assigned to receive every 2 month 4 mL of intramuscular ISG from unscreened donors (450 partners) or placebo (499 partners). Seven partners developed acute HCV infection (increased aminotransferase levels and appearance of HCV-RNA): six of the placebo group (incidence density [ID] 12.00/1,000 person year; 95% confidence interval [CI] 3.0 to 21.61), and only one of the ISG-treated group (ID 1.98/1,000 person year; 95% CI 0 to 5.86). The risk of infection was significantly higher in controls versus treated individuals (p = 0.03). Six couples had genotype 1b (85%), and one couple had genotype 1a; HCV sequence homology strongly supported sexual transmission. Our trial demonstrates that HCV infection can be sexually transmitted and quantifies the risk of sexual transmission: for every year of at-risk sexual relationship, almost 1% of the partners became infected. Intramuscular ISG is safe and well tolerated. Unlike ISG from screened donors, ISG from donors unscreened for anti-HCV contains high titers of anti-gpE1/gpE2 neutralizing antibodies and high neutralizing activity. Anti-HCV hyperimmune globulin could be prepared from anti-HCV-positive blood units and could be used to protect sexual partners and in other at-risk situations of exposure to HCV infection.