In forensic science one frequently has to deal with unidentified skeletonised remains. When conventional methods of identification have proven unsuccessful, forensic facial reconstruction (FFR) may be used, often as a last resort, to assist the process. FFR relies on the relationships between the facial features, subcutaneous soft tissues and underlying bony structure of the skull. The aim of this study was to develop soft tissue thickness (STT) values for South African black females for application to FFR, to compare these values to existing literature or databases, and to test the accuracy and recognisability of reconstructions using these values. It also established whether population-specific STT values are necessary for FRR. Computerised tomography scanning was used to determine average populationspecific STT values at 28 facial landmarks of 154 black females. The Manchester method of facial reconstruction was employed to build faces, for which antemortem photographs were available, on two skulls that were provided by the South African Police Service’s (SAPS) Forensic Science Laboratory. Different data sets of STT values, namely values from this study, two sets of data from American blacks and a South African mixed ancestry group, were used to build four faces for each of the skulls. Two identification sessions were then held. In the first session, 30 observers were asked to select matches from a random group of 20 photographs of black females which included the two actual images. The identification rates calculated for each photograph revealed that the highest rates of a positive match were for the reconstructions based on South African values. In the second session another group of 30 volunteers were asked to match to each photograph the most similar of the four reconstructions made of that particular individual. The reconstructions with STT values from the current (South African) study were selected more often than the other data sets. Although shortcomings do exist, the identification sessions indicated that FFR can be of value. Furthermore, population-specific STT values are important, since skulls reconstructed using these values were selected or identified statistically significantly more often than the others. AFRIKAANS : In forensiese wetenskap het mens dikwels te doen met ongeïdentifiseerde skeletmateriaal. Wanneer die konvensionele metodes van identifikasie onsuksesvol is, mag forensiese gesigsrekonstruksie (FGR) gebruik word, dikwels as `n laaste uitweg, om die proses te help. FGR is afhanklik van die verhouding tussen die gelaatstrekke, subkutane sagte weefsels en onderliggende benige struktuur van die skedel. Die doel van hierdie studie was om sagte weefsel dikte (SWD) waardes vir Suid-Afrikaanse swart vroue te ontwikkel vir gebruik met FGR, om hierdie waardes te vergelyk met bestaande literatuur of databasisse, en die akkuraatheid en herkenbaarheid van rekonstruksies waar hierdie waardes gebruik was te toets. Dit is gedoen ten einde vas te stel of bevolking-spesifieke SWD waardes nodig is vir FGR. Gerekenariseerde tomografie skandering is gebruik om die gemiddelde bevolkingspesifieke SWD waardes op 28 gesigslandmerke van 154 swart vroue te bepaal. Die Manchester metode van gesigsrekonstruksie is gebruik om twee skedels, waarvan antemortem foto’s beskikbaar was en wat voorsien is deur die Suid Afrikaanse Polisie Diens (SAPD) se Forensiese Wetenskap Laboratorium, op te bou. Verskeie data stelle vir SWD waardes, naamlik waardes verkry in hierdie studie, twee stelle Amerikaanse waardes vir swart vroue en `n Suid Afrikaanse groep van gemengde afkoms, is vir hierdie studie gebruik om vier gesigte van elk van die skedels te bou. Twee identifikasie sessies is gehou. In die eerste sessie is 30 deelnemers gevra om passende foto’s uit `n algemene versameling van 20 foto’s van swart vroue te kies. Dit het die twee ware gesigte ingesluit. Die identifikasie waardes wat bereken is vir elke foto het getoon dat die hoogste waardes vir die werklike foto’s verkry is op rekonstruksies gebasseer op Suid-Afrikaanse waardes. In die tweede sessie was `n ander groep van 30 vrywillgers gevra om die mees soortgelyke van die vier rekonstruksies by die foto van die betrokke individu te pas. Die rekonstruksies met SWD waardes van die huidige (Suid Afrikaanse) studie was meer dikwels gekies as die van ander data stelle. Hoewel verskeie tekortkominge bestaan, het die identifikasie sessies getoon dat FGR van waarde kan wees. Verder is bevolking-spesifieke SWD waardes belangrik, aangesien skedels wat opgebou is met hierdie waardes statisties beduidend meer dikwels gekies of geïdentifiseer is as die ander. / Dissertation (MSc)--University of Pretoria, 2010. / Anatomy / unrestricted