A hepatitis immunization field trial, using a Chinese hamster ovary (CHO) cell recombinant vaccine, was implemented for newborns in the Austral archipelago of French Polynesia in 1988. Three different schedules were used: (1) four vaccine doses at months (M) M0, M1, M2 and M12; (2) three vaccine doses at M0, M1 and M6; and (3) three vaccine doses at M0, M1 and M12. The programme evaluation was performed yearly at fixed dates, i.e. October-November 1989, 1990 and 1991. After the third year, of the 582 children who received one or more doses of vaccine, four were HBsAg carriers. After one or two doses, 88 and 98%, respectively, had seroconverted for at least one of the two measured antibodies, anti-HBs or anti pre-S2. After three doses, seroconversion rates and geometric mean anti-HBs titres were, respectively, 94% and 187 mIU ml-1 using schedule M0, M1, M2; 95% and 507 mIU ml-1 using schedule M0, M1, M6 and 96% and 476 mIU ml-1 using schedule M0, M1, M12. After four doses (M0, M1, M2, M12) the corresponding results were 99% and 1518 mIU ml-1. One of the 16 vaccinated neonates born to HBsAg/HBeAg-positive mothers was an HBsAg carrier, implying a protective rate for the prevention of perinatal transmission of 93%. Overall, these results indicate that, in field conditions, indiscriminate vaccination of newborns with a CHO-recombinant vaccine without hepatitis B immunoglobulin (HiBG) resulted in high immunogenicity. Final evaluation in 1993-1994 will permit confirmation of the effectiveness of the two three-dose vaccine schedules.