A surgical luminaire containing two types of phosphor-converted LEDs was employed to illuminate a large midline incision in a 90 kg pig. The PC-LEDs were proportionally blended to create four spectra defined by their approximate correlated colour temperatures: 3000, 4000, 4500 and 5100 K. Sixteen surgeons evaluated the colour appearance of internal anatomy under each spectrum. All four spectra were rated as highly effective. There was no significant difference between the 4000, 4500 and 5100 K spectra. The 3000 K spectrum was rated as less effective, though this does not mean it was ineffective. The spectrum by participant interaction was significant where it could be estimated, suggesting that surgeons may develop individual opinions about which spectra are most effective. While the surgeons moderately agreed that the colour of internal anatomy was altered by the different spectra, they were undecided whether the differences would increase the risk of surgical error or influence surgical decisions. They were also largely undecided whether they would select different spectra during a single procedure, for different procedures, or for different anatomy. Overall, the linear mixing for variable CCT employed in this experiment did not provide obvious utility for surgeons making visual evaluations.