The colour–magnitude diagram (CMD) of NGC 1851 presents two subgiant branches (SGBs), probably due to the presence of two populations differing in total C+N+O (carbon+nitrogen+oxygen) content. We test the idea that a difference in total C+N+O may simulate an age difference when comparing the CMD of clusters to derive relative ages. We compare NGC 1851 with NGC 6121 (M4), a cluster of very similar [Fe/H]. We find that, with a suitable shift of the CMDs that brings the two red horizontal branches at the same magnitude level, the unevolved main sequence and red giant branch match, but the SGB of NGC 6121 and its red giant branch ‘bump’ are fainter than in NGC 1851. In particular, the SGB of NGC 6121 is even slightly fainter than the faint SGB in NGC 1851. Both these features can be explained if the total C+N+O in NGC 6121 is larger than that in NGC 1851, even if the two clusters are coeval. We conclude by warning that different initial C+N+O abundances between two clusters, otherwise similar in metallicity and age, may lead to differences in the turnoff morphology that can be easily attributed to an age difference.