Epidemiological investigations of outbreaks of infectious disease have
to be carried out rapidly but must be methodologically sound. In 14 o
f 25 consecutive Salmonella outbreaks in Wales from 1986 to 1990 case-
control or cohort studies were undertaken. Food vehicles of infection
were identified by statistically significant associations with cases i
n 13 of the 14, including three outbreaks where there were less than 1
0 cases. The particular problems of epidemiological field investigatio
ns are discussed.