OBJECTIVES AND METHODS: With the purpose of better understanding the efficacy of the lower titer [4 x 10(4) plaque-forming units (pfu)] tetravalent rhesus-human reassortant rotavirus vaccine (RRV-TV) against diarrheal episodes of different severities, the Peruvian and Brazilian efficacy data were reanalyzed with a 20-point scoring system. Mild, moderate/severe and very severe rotavirus diarrhea were scored as 0 to 8, 9 to 14 and >14, respectively. RESULTS: In the Peruvian study one dose of vaccine yielded 64% (P = 0.04) protection against pure cases of rotavirus disease (i.e. those in which no other enteropathogen was found) with clinical scores ranging from 9 to 14. Protective efficacy against very severe rotavirus gastroenteritis could not be assessed because of the small number of cases. In Brazil there was a trend in preventing "all" and "pure" cases of rotavirus diarrhea scored 9 to 14 (44%, P = 0.06, and 45%, P = 0.08, respectively) and the vaccine was 75% (P = 0.02) protective against pure rotavirus diarrhea scored >14. No protection was observed for mild rotavirus diarrhea (scores <9). These data were compared with those from trials in Venezuela (4 x 10(5) pfu/dose), US (4 x 10(4) pfu/dose and 4 x 10(5) pfu/dose) and Finland (4 x 10(5) pfu/dose). Combining the Peruvian (one dose, pure cases) and Brazilian studies together, the levels of protection against 9- to 14-scored rotavirus diarrhea are comparable with those from the Venezuelan (47%) and American (57, 57 and 65%) efficacy trials. In Brazil the level of protection (75%) against pure, >14-scored rotavirus diarrhea is similar to the efficacy rates yielded in the three US trials (82, 80 and 69%) and the Finnish trial (100%) for episodes of the same severity. CONCLUSIONS: Our reanalysis provides evidence that, at least against moderate/severe rotavirus gastroenteritis, RRV-TV, 4 x 10(4) pfu/dose is potentially as efficacious as RRV-TV, 4 x 10(5) pfu/dose, even in settings with very high rotavirus disease burden. The reanalysis of the Peruvian data suggests that one and three vaccine doses may yield similar efficacy rates. It is also suggested that vaccine efficacy against most severe episodes in Peru and Brazil was not evident because of the trial design used in those studies (i.e. prospective, active home surveillance rather than a catchment trial), resulting in too few cases of severe disease even in the placebo group. To confirm these findings, future trials with this vaccine are necessary in developing countries with high diarrhea morbidity rates. These trials should use catchment designs and focus on the evaluation of the efficacy of one or three doses of RRV-TV against moderate to severe/very severe rotavirus diarrhea.