In 1997, 403 babies in England and Wales died suddenly, unexpectedly, and for no reason (despite a thorough postmortem examination)—a rate of 0.63 per 1000 live births (press release from the Foundation for the Study of Sudden Infant Death, August 1998). It is salutary to recall that 10 years ago the rate of sudden infant death syndrome (SIDS) was 2.3 per 1000 live births.1 Much of this remarkable improvement, a trend shared by many other Western countries, is attributed to the simple four point advice included in the “reduce the risk” campaign first launched in Britain in November 1991, in particular, the importance of avoiding the prone sleeping posture (although it is worthy of note that the incidence of SIDS had begun to fall a few years before this). This success story should however, not be allowed to conceal the continuing tragedy of about nine babies each day who are victims of SIDS. This residuum (still constituting the largest single group of deaths in infants 1 week to 1 year old) is stubbornly resisting attempts to shift it—the incidence of SIDS has not changed significantly since 1995.

Most of these deaths are in families where mothers smoke during pregnancy and where there is household exposure to tobacco smoke after birth.2 Some view this association a direct one2 3: exposure to products of cigarette smoke before and after birth perhaps interfering with lung growth or neural mechanisms involved in the control of breathing. Stopping parents smoking, it has been estimated, could reduce the risk of SIDS by almost two thirds.2 But it is also important to consider the additional possibility that smoking is a marker for other, as yet unknown, pathophysiological mechanisms leading to death, with which adverse social circumstances (particularly the illegal use of drugs, …