Sudden infant death syndrome (SIDS), as a diagnosis of exclusion, cannot, by definition, be “diagnosed” with absolute certainty. Attempts have been made by various authors to clarify rather than modify Beckwith’s initial 1969 definition,1 but all have included the inescapable fact that the investigation “fails to demonstrate an adequate cause of death”. SIDS is therefore the label used when we don’t know why the baby died. After almost 40 years of intensive study it is also becoming increasingly unlikely that one single causal mechanism would explain these deaths. Thus trying to calculate a recurrence risk for deaths due to disparate and largely unknown causes, potentially involving several independent mechanisms, is a difficult task. The hypothesised pathophysiology in many SIDS deaths is complex. For example, the presence of certain common polymorphisms of immune function (eg, affecting interleukin (IL) 6, IL10, IL1β) may predispose to unexpected death in the presence of an unrecognised or mild infection. Although genetic factors are crucial in such deaths, the presumed interactions between the genetic polymorphisms and infection at a vulnerable stage of development (see the “triple risk hypothesis” below) means that there is probably a relatively low risk of recurrence in future children. In other deaths limited pathological investigations or failure to recognise and further investigate suggestive histological findings could result in the failure to identify a genetically determined metabolic disorder (eg, of fat oxidation) with a relatively high risk of recurrence and of death in future siblings. The frequency with which cardiac chaelopathies (eg, prolonged QT syndrome) contribute to unexpected deaths in infancy seems to vary widely between studies. The large number of genetic variants that can give rise to such conditions …