The evaluation of scar colour is, at present, usually limited to an assessment according to a scar assessment scale. Although useful, these assessment scales only evaluate subjectively the degree of scar colour. In this study, the reliability of the subjective assessment of scar colour by observers is compared to the reliability of the measurements of two objective colour measurement instruments. Four independent observers subjectively assessed the vascularisation and pigmentation of 49 scar areas in 20 patients. The degree of vascularisation and pigmentation was scored according to a scale ranging from '1', when it appeared to be like healthy skin, to '10', which corresponds to the worst imaginable outcome of vascularisation or pigmentation. The observers also scored the pigmentation categories of the scar (hypopigmention, hyperpigmention or mixed pigmentation). Finally, each observer measured the scar areas with a tristimulus colorimeter (Minolta Chromameter®) and a narrow-band simple reflectance meter (DermaSpectrometer®). A single observer could reliably carry out measurements of the DermaSpectrometer and the Minolta Chromameter for the evaluation of scar colour (r=0.72). The vascularisation of scars could also be assessed reliably with a single observer (r=0.76) whereas for a reliable assessment of pigmentation at least three observers were necessary (r≥0.77). The agreement between the observers for the pigmentation categories also turned out to be unacceptably low (k=0.349). This study shows that an overall evaluation of scar colour with the DermaSpectrometer and the Minolta Chromameter is more reliable than the evaluation of scar colour with observers. Of both instruments for measuring scar colour, we prefer, because of its feasibility, the DermaSpectrometer.