Thesis (MSc (Chemistry)) -- University of Limpopo, 2011. / Surveys indicate that most South African origin women want long and straight hair (Mamabolo and Summers, 2006). Such women will therefore use a relaxer to straighten their hair. Many such preparations cause damage to the hair and scalp; hence identification of an effective and mild relaxer would be advantageous. This dissertation investigates South African origin hair structure and goes on to describe the effects of relaxers on South African origin hair. This work also analyses the effects of the lye and no-lye relaxers separately. The work was divided into two sections. The first section was the clinical study where two types of relaxers (‘lye’, a sodium hydroxide base relaxer and ‘no-lye’, a guanidine hydroxide base relaxer) were applied on the new outgrowth representing natural hair in a half-head study design of five South African origin female volunteers. Subjective (subject self-assessment) and objective (Researcher assessment) hair assessments were performed. The second section was the analysis of the hair samples by Reversed-Phase High Performance Liquid Chromatography (HPLC) using the Pico Tag as well as Electron Microscopy. There was no erythema detected in the clinical study on the scalp of any of the subjects post-application. Both the researcher and the subjects assessed the performance of the no-lye relaxer to be better than the lye relaxer in terms of straightness. The researcher also assessed the performance of the no-lye relaxer to be better than the lye relaxer in terms of softness, shininess, volume and dryness. The performance of the two types of relaxers was assessed to be the same by both the researcher and the subjects in terms of length and damage. There was a statistically significant (p< 0.1) decrease in the amount (g/100 g hair) of cystine after treatment compared to before treatment with both the lye relaxer (median [range]) (7.8 [2.5-9.9] vs. 9.1 [6.7-11.9]; p= 0.086) and the no-lye relaxer (4.0 [2.9-4.8] vs. 9.1 [6.7-11.9]; p= 0.005); this decrease was significantly greater (p= 0.086) for the no-lye relaxer. There was also a statistically significant decrease in the amount of lysine after treatment compared to before treatment with both the lye relaxer (2.0 [1.5-2.3] vs. 2.1 [2.0-2.6]; p= 0.082) and the no-lye relaxer (2.0 [1.5-2.2]; p= 0.036); this decrease was not significantly different (p= 0.920) for the two types of relaxer. No significant differences were found in the levels of the remaining 15 amino acids analysed. For all the subjects no physical evidence of hair damage was observable from the scanning electron microscopy images of the hair shafts and the cuticles. The longer wash-off time confirmed the claimed relative safety of ‘no-lye’ relaxers. The no-lye relaxer performed better overall than the lye relaxer in terms of the hair quality parameters assessed. A decrease in cystine levels is consistent with better performance in terms of hair straightness. The results from electron microscopy were not conclusive.