BACKGROUND: The "Back to Sleep" campaign in 1991 resulted in a dramatic decrease in the incidence of sudden infant death syndrome (SIDS). The proportion of presumed SIDS deaths being actually suspicious deaths from airway obstruction is likely to have become relatively greater. There is usually little pathological evidence to suggest smothering, but intra-alveolar haemorrhage appears to be more prominent in cases where interference with the airway is suspected. AIM: To attempt to quantify intra-alveolar haemorrhage to see whether it could be used as a marker to distinguish between smothering/overlaying and SIDS. METHODS: Intra-alveolar haemorrhage was quantified using digital image analysis on haematoxylin/eosin stained sections taken from the lungs of 62 consecutive infants who had died suddenly and unexpectedly. Cases were initially classified according to the original cause of death. After quantitation, the case histories were critically reviewed. Three pathologists independently made microscopic assessments of the degree of intra-alveolar haemorrhage in the first 24 cases to see whether these accurately reflected the quantitative results. RESULTS: 73% of those infants with a history suggesting possible involuntary overlaying and 45% of those with a history suspicious of smothering had significant intra-alveolar haemorrhage (> 5% of total lung surface area assessed). From the history, the cause of death in 11 cases initially classified as SIDS would better have been given as "Unascertained." Simple microscopic assessments underestimated the true extent of the haemorrhage in 33% (8/24). CONCLUSIONS: If a moderate degree (at least 5%) of pulmonary parenchymal haemorrhage is observed, this may be an indicator of airway obstruction for a significant period, either from overlaying or possibly smothering. The diagnosis of SIDS may be being used inappropriately in such cases.