The authors report on a community-based, nonrandomized, open-label study, conducted during the 2000-2001 influenza season in Temple-Belton, Texas, of the protective effectiveness of trivalent, cold-adapted, influenza virus vaccine (CAIV-T) in children aged 18 months-18 years. The dominant circulating strains in 2000-2001 were influenza A/New Caledonia/20/99 (H1N1) and influenza B/Sichuan/379/99. Children had access to CAIV-T during the 1998-1999, 1999-2000, and 2000-2001 influenza seasons. The vaccine included influenza A/Sydney/5/97 (H3N2) and B/Beijing/184/93-like (B/Ann Arbor/l/94) strains in all three seasons. The vaccine included A/Beijing/262/95 (H1N1) in 1998-1999 and 1999-2000, which was replaced by A/New Caledonia/20/99 (H1N1) in 2000-2001. When medically attended acute respiratory illness (MAARI) was used as the outcome, the protective effectiveness for children vaccinated in 2000 was 18% (95% confidence interval (CI): 11, 25). Based on a combination of a validation sample of surveillance cultures and the MAARI outcome, protective efficacy against combined influenza A (H1N1) and B was 79% (95% CI: 51, 91). The efficacy estimate, after accounting for missing influenza culture status, against influenza A (H1N1) alone was 92% (95% CI: 42, 99) and against a new variant of influenza B alone was 66% (95% CI: 9, 87). CAIV-T provides substantial protection against a mixture of influenza A (H1N1) and B. Results demonstrate the powerful potential of using validation sets for outcomes in vaccine field studies.