The 202 deaths connected with burns in 1976-1986 included 134 in which death occurred during the fire. In 49 cases the face was so badly charred that it was not possible to judge whether hemorrhage had occurred in the eyelids and/or conjuctivae. Such blood extravasation was seen in 14 of the remaining 85 cases, taking the form of petechiae in some cases and in others the form associated with more extensive extravasation, in some cases hemolytic. No other features known to cause such hemorrhages were present, so that they are attributed to the burning process itself. Signs of extensive pronounced burns were found in 13 of these cases, and it seemed that high-degree burns on the neck and trunk with less severe burns on the head were particularly likely to be associated with such hemorrhages. In most cases the flames had quickly come into direct contact with the body (clothing, bed or chair had caught fire), leading to rapid death, as documented by the fact that usually very little or no soot at all had been aspirated and by the low level of carbon monoxide intoxication (max. COHb 22%). One plausible explanation for the development of hemorrhage is the supposition first expressed over 20 years ago that when the circulation is maintained rapid burn-induced shrinkage of the skin of the neck has the effect of strangulating the victim; hemorrhage of eyelids and/or conjunctivae in such cases be could regarded as a vital reaction to the effect of fire. This is particularly significant, insofar as evidence of hemorrhage of this kind was found predominantly in bodies in which other vital signs indicative of the effects of burning were sparse and slight or even totally absent. In addition, specific examination of the laryngeal area revealed congestion-induced extravasation at various points, as well as petechial hemorrhage in the mucous membrane.