Clustering of malformations and especially of spina bifida in Sweden was studied using the Register of Malformations and the Medical Birth Records Register. The incentive to this study was provided by the observation of a cluster of malformations, notably spina bifida, and perinatal deaths, locally thought to be related to the use of herbicides in forestry. This cluster was verified but it was shown that a high rate of spina bifida existed in the region before the introduction of herbicides in forestry. Some other clusters of spina bifida were identified and also a number of clusters of other malformations. Case-history studies were performed but revealed no lead to an explanation. Principles in the origin of the clusters and their analysis are discussed.