Sweden has experienced an epidemic of symptomatic coeliac disease that has no likeness anywhere else in the world. This is quite unique for a disease that is genetically dependent, immune-mediated and chronic, and suggests an abrupt increase and decrease, respectively, of one or a few causal factors influencing a large proportion of Swedish infants during the period in question. We have shown that half of the epidemic was explained by an increase in the proportion of infants introduced to gluten in comparatively large amounts after breast-feeding had been ended. This was partly an effect of societal changes in national dietary recommendations and the food content of industrially produced infant foods. Thus, these changes over time in infant feeding practices had a large public health impact. In fact, when the Swedish epidemic began, the increase in incidence rate was larger in girls than in boys, as was the decrease during the post-epidemic period. Moreover, children born during summer had an increased risk for coeliac disease, possibly as they were mostly introduced to dietary gluten during winter when infections are more common. Notably, birth cohorts of the epidemic and post-epidemic periods differ considerably regarding coeliac disease occurrence at comparable ages, even when followed up to school age. A longer follow-up will reveal to what extent new cases develop later in life, and to what extent this difference in cumulative incidence remains. However, mass screening studies of both the epidemic and post-epidemic cohorts at comparable ages are also planned to determine to what extent 'silent' disease cases develop. Continuing to explore the Swedish epidemic of coeliac disease by means of an epidemiological approach provides a unique opportunity. This may include increasing our understanding of what determines the clinical expression of the disease, exploring the potentially causal role of environmental exposures, and possibly also identifying strategies for primary prevention.