In this issue of the journal the possibility that heme oxygenase activity is responsible for the delay in the appearance of the yellow color in bruises has been investigated [1]. Although bruises may have been a ‘‘very much neglected branch of injuries’’ [2] for many years, recent investigations, including the current paper [1] have revealed some unexpected and highly relevant findings. Bruises in the skin and subcutaneous tissues occur when there has been extravasation of blood following the application of blunt force. Generally the requirements for their formation are injury, compromise of the integrity of vessel walls, and blood pressure [3]. The most useful bruises in a forensic context are those that have a patterned appearance reflecting the nature of the impacting object. Questions that are commonly asked in court regarding bruises involve their age, the amount of force required for their production, and whether they occurred before or after death. Although it was once considered possible to provide answers to all of these queries, recent studies have shown that forensic assessments, particularly of the age of these injuries, may be less reliable than was previously considered. For example, it was once generally accepted that bruises could be dated from their macroscopic appearance and standard text books gave quite precise times for this, with bruises appearing red, purple, and swollen in the first 2 days, green from 4 to 7 days, yellow from 7 to 10 days, with resolution between 14 and 30 days. Of note, even then there was disagreement among the texts in the exact timeline for these color changes [4–7]. The paper by Langlois and Gresham in 1990 debunked this certainty, and demonstrated that the only reliable comment that could be made was that yellow bruises were likely to be older than 18 h [8]. It should be noted, however, that this study did rely on photographs of bruises, which may present additional complication in interpretation (see below). The reasons for the great variation in individual bruising responses to blunt trauma are varied and include differences in the nature of the impacting object and forces, age of the victim, variations in subcutaneous tissue thickness, variability in bleeding tendencies, the effects of underlying disease states, the influence of medications (both pharmaceutical and herbal) [9], and background skin pigmentation. Other issues concern considerable interobserver variability in the perception of yellow in bruises [10] and in the identification of specific colors [11], i.e. one observer’s brown may be another observer’s yellow, and the tendency for single bruises from one episode of trauma to develop different colors in different areas at the same time. All of this tells us that dating of bruises macroscopically is, at best, highly inaccurate. Photographing bruises introduces further variables as a bruise can be made to disappear if a picture is overexposed; in addition, an apparent ‘‘bruise’’ can be created if shadows occur from under-exposure. For this reason photographs should not be relied upon for the assessment of the age of bruises. Ideally bruises should be photographed in color at autopsy in suspicious cases, both before and after incision (the latter to confirm the presence of subcutaneous hemorrhage). Photography in black and white using infra-red or alternative light source illumination may provide additional information, and trace wound patterns may be detected by use of an alternative light source (although alternative light source [polilight ] illumination with digital image analysis does not assist in determining age). R. W. Byard (&) N. E. I. Langlois Discipline of Anatomy and Pathology, Level 3 Medical School North Building, The University of Adelaide, Frome Road, Adelaide, SA 5005, Australia e-mail: roger.byard@sa.gov.au