To design a safe and effective synthetic peptide vaccine against rubella virus (RV) infection, it is necessary to identify immunodominant T-cell epitopes of RV structural proteins. To define such epitopes, 49 overlapping synthetic peptides (17-34 residues in length) corresponding to more than 95% of the amino acid sequence of RV virion proteins E1 (23 peptides) and C (11 peptides) and all of E2 (15 peptides) were synthesized and tested for their capacities to induce proliferative responses of rubella-specific T-cell lines and T-cell clones derived from 4 study groups (5 women infected with RV in pregnancy, 5 patients with congenital rubella syndrome, 5 seropositive healthy donors, and 5 RV vaccine recipients). The most frequently recognized epitopes were E1-21 (residues 358-377) with 11/20 responders, E2-4 (residues 54-74) with 6/20 responders, and C11 (residues 255-280) with 11/20 responders, respectively. E1-10 (residues 174-193), E1-16 (residues 272-291) and E1-18 (residues 307-326) were responded to strongly by corresponding T-cell clones, and were recognized by 4 or 5 T-cell lines. T-cell lines derived from three congenital rubella syndrome patients did not respond to any of the synthetic peptides. The results showed that more T-cell epitopes were present in E1 (19/23) and C (10/11) than in E2 (8/15). The identification of T cell sites recognized frequently by RV-infected or -immunized populations could provide the basis for selecting candidate T-cell epitopes for the development of an effective synthetic vaccine against rubella.