Objects: In July 2 2010, a diarrhea outbreak occurred among the workers in a company in Gyeungju city, Korea. An epidemiological investigation was performed to clarify the cause and transmission route of the outbreak. Methods: We conducted a questionnaire survey among 193 persons, and we examined 21 rectal swabs and 6 environmental specimens. We also delegated the Daegu Bukgu public health center to examine 3 food service employees and 5 environmental specimens from the P buffet which served a buffet on June 30. The patient case was defined as a worker of L Corporation and who participated in the company meal service and who had diarrhea more than one time. We also collected the underground water filter of the company on July 23. Results: The attack rate of diarrhea among the employees was 20.3%. The epidemic curve showed that a single exposure peaked on July 1. The relative risk of attendance and non-attendance by date was highest for the lunch of June 30 (35.62; 95% CI, 2.25 to 574.79). There was no specific food that was statistically regarded as the source of the outbreak. $Bacillus$ $cereus$ was cultured from two of the rectal swabs, two of the preserved foods and the underground water filter. We thought the exposure date was lunch of June 30 according the latency period of $B.$ $cereus$. Conclusions: We concluded the route of transmission was infection of dishes, spoons and chopsticks in the lunch buffet of June 30 by the underground water. At the lunch buffet, 50 dishes, 40 spoons, and chopsticks were served as cleaned and wiped with a dishcloth. We thought the underground water contaminated the dishes, spoons, chopsticks and the dishcloth. Those contaminated materials became the cause of this outbreak.