Recent studies of blue tits, Parus caeruleus, have found sexual selection and a viability-indicating function of the structural ultraviolet and blue crown plumage, but the reasons for this signal variation are not understood. Furthermore, studies in England and Sweden have yielded somewhat different results (particularly with regard to the spectral position of the reflectance peak). Here we investigate whether the blue tit UV/blue ornament varies with time of year since such variation might be relevant to the signalling function as well as the apparent difference between populations. From 400 blue tits captured at two different localities in Sweden, we found that objective measures of ‘hue’ (spectral location), ‘chroma’ (spectral purity) and ‘brightness’ (spectral intensity), varied substantially with season. Just after moult (October), crown ‘hue’ is maximally UV-shifted (359 nm for males and 373 nm for females). Thereafter the peak drifts upwards and by the time of nestling feeding (June) male reflectance peaks at 404 nm and female at 413 nm. This change is probably due to feather wear as well as fat and dirt accumulation, which might constitute an additional male quality cue. Our results suggest that it is important to consider plumage age when exploring variation in structural plumage coloration, and that it can largely explain the difference between the British and Swedish studies.