The medical symptom 'iceberg' and 'trivia' were defined in terms of people's own perceptions of their symptoms and their subsequent referral behaviour. The data were collected by household interviews of patients registered at a health centre and included information on personal and environmental characteristics. Bivariate and multivariate analysis was used to explore associations between those who were part of the symptom 'iceberg' or 'trivia', and factors which might have caused such incongruous referral behaviour.