W.A. Thornton [W. A. Thornton, Toward a more accurate and extensible colorimetry, Parts I–III, Color Res. Appl. 17, 79–122, 162–186, 240–261 (1992).] reported significant discrepancies in Maxwell colour matches when using different sets of primary lights. His experiments were performed by a small, mainly older group of observers. the aim of this investigation was to test these findings over wider age ranges. We matched a broad‐band white light of constant luminance with two sets of three monochromatic primaries. These sets were similar to Thornton's AP set (497, 579, and 653 nm) and Thornton's PC set (452, 533, and 607 nm). Thirty‐two observers performed the tests with 10° fields. the chromaticities of the matches made with the AP set had a large variance and were centered around the chromaticity of the reference white. the chromaticities of the PC‐primary matches were clustered closely together and were centered at a point significantly different from the chromaticity of the reference white. the magnitudes, but not the directions, of the dispersions of the matching chromaticities were consistent with age related changes in the spectral transmission of the lens.