A rotavirus vaccine is likely to be licensed in the near future but the cost-effectiveness of the vaccine remains an issue. Recent advances in the outpatient management of diarrhea and the use of short (<24 hour) stay care may have altered the cost of medical care for rotavirus. To examine the impact of rotavirus infection prior to the licensure of a rotavirus vaccine, we conducted active rotavirus surveillance from March 16-September 17, 1997 on all hospitalized children <5 years of age with signs and symptoms of acute gastroenteritis (AGE). Stool samples were tested by ELISA for rotavirus antigen. Positive specimens were then serotyped using type-specific monoclonal antibodies directed against G-serotypes 1, 2, 3 and 4. Of the 148 children admitted with AGE, 39.2% were admitted to the short-stay unit and 15% were<6 months of age. Overall, 84.5% of AGE patients shed rotavirus. Short-stay patients were more likely to have rotavirus compared to inpatients (94.8% vs. 77.8%; p=.005). Children ≥.6 months of age were more likely to have rotavirus than those <6 months (88.9% vs. 59.1%;p<.001). Of the 125 specimens available for typing, 92% were typable. Of those infections with a known serotype, G1 was the predominant serotype accounting for 72.2% of infections, but unlike previous evaluations in Cincinnati, all 4 serotypes were identified [G4 (12.2%), G3 (10.4%), G2 (5.2%)]. There was no significant difference of the mean age by serotype. Rotavirus still accounts for a large proportion of cases of severe AGE, many of which are admitted to short-stay units which would be missed in hospital discharge surveys. Each of the four major serotypes were found during one season, indicating a need for a vaccine which provides protection against multiple serotypes.