Norwalk virus has been implicated in shipboard diarrheal disease outbreaks throughout Asia. A large outbreak of suspected Norwalk virus was investigated on a U.S. Naval aircraft carrier following the clinical recognition of 450 cases of gastroenteritis over a 2-week period (September 14-28, 1997) during coastal exercises. A random sampling of 44 cases from 450 personnel who sought medical attention was compared with 19 controls. Junior enlisted sailors and marines comprised 97% of all cases. There was no evidence of shipboard geographic clustering of cases. Furthermore, no single food type was associated with illness on the basis of comparative analysis (cases versus controls). Principal case signs and symptoms reported included watery stools (89%), nausea (82%), and vomiting (77%). Anecdotal reports indicated > 50% of the cases received rehydration therapy. An absence of fever was also noted in 32% of the cases and only 5% had blood in their stools. The mean duration of illness was 37 hr, with a range of 3-96 hr. Laboratory findings based on reverse transcription-polymerase chain reaction and Southern hybridization methods showed that 21 (72%) of 29 patients had evidence of the UK2 prototype of the Norwalk virus. A cross-sectional study of 131 crew members from the ships population (n = 4,200) showed an attack rate of 44%. Attack rate is a variant of an incident rate applied to a narrowly defined population observed for a limited period of time, such as during an outbreak. The numerator is people who get sick and the denominator is people (population) at risk. An extrapolation of these findings suggests as many as 1,806 sailors may have been affected during the outbreak, of which only 26% (of the 57 outbreak related cases) where identified from sick call records. There was no difference in the mean ages between outbreak and non-outbreak affected crewmen, or geographic clustering based on berthing or work spaces. Outbreak-related cases reported signs and symptoms of watery-stools (79%), nausea (65%), and vomiting (47%). The mean duration of illness was 28 hr, ranging from 2 to 96 hr. Thirty-one percent of outbreak affected cases reported a sick call visit. Loss of work was reported by 39% of the outbreak affected population. This report documents the epidemic potential of Norwalk virus and the associated impact on fleet operational readiness. Additionally, that this outbreak occurred against a background of 3 other consecutive gastroenteritis outbreaks onboard the same ship (March 1997, February/March 1998, and June 1998), all sharing the same clinical and epidemiologic profiles, suggests possible shipboard persistence of Norwalk virus over time, despite periodic ship-wide disinfection efforts.