Published reports of clusters of congenital anomalies, traditionally define
d as an aggregation in time and/or space of malformed cases, are interestin
g to analyse in terms of emergence, management and initiator of a public he
alth decision. Through some examples of clusters for which a suspected sour
ce has been suggested, for those having identified important causal relatio
nship or for those where the cluster appeared without any explanation, the
paper shows the different steps that were taken after the initial 'alarm' a
nd the time spent between the alarm and a final conclusion. If basic keys s
uch as accurate field investigation, reliable estimation of the expected nu
mber and etiological evaluation, are the rule, the handling should remain f
lexible to take into account the particularity of each cluster. Among the c
lusters of congenital anomalies published in the last 20 years, very few of
them were clearly explained.