The ecology of non-O 1 Vibrio cholerae and Vibrio mimicus as causes of cholera-like diarrhea or seafood-associated gastroenteritis has been investigated in Toyama Prefecture since 1980. The relationship between biological or serological characteristics of the isolates and their enteropathogenicity is discussed. Overall isolation rates from river water, sea water, and fish were 24.0, 59.5, and 33.7%, respectively, the isolation frequency being, in general, extremely high in the summer season, although the organisms were detected all year around in the case of sea water. Most isolates from river water were unable to grow on plates of TCBS agar to which colistin was added at a concentration of 1 microgram/ml (CL-TCBS). These strains quickly fermented cellobiose. O-51 and O-70 were the two most frequently detected serogroups among them and they did not show enteropathogenicity in the rabbit ileal loop ( RIL ) test. On the other hand, almost all isolates from sea water and fish as well as those from human diarrhea cases were able to grow on CL-TCBS, but were unable to ferment cellobiose quickly. O-36, O-10, O-6, O-8, O-39, and O-26 were the dominant serogroups of these isolates, and some of them showed enteropathogenicity in the RIL test. Six out of 98 isolates from river water, 14 out of 116 from sea water, and 19 out of 112 from fish were classified as Vibrio mimicus . All of these strains were able to grow on CL-TCBS and quickly fermented mannose but not cellobiose. O-41 was the most common serogroup among them and some of these strains showed enteropathogenicity in the RIL test. Production of a cholera-like enterotoxin among the isolates in Toyama Prefecture, if any, seemed to be poor.