ROWAN-ROBINSON has presented evidence that radio galaxies exhibit remarkably strong evolutionary effects with cosmological epoch1. Examination of his evidence, however, indicates that, although an evolutionary effect may indeed be present, no reliable estimate of the time scale involved can yet be made in the way he proposes. This is because (i) the mean value of ν/νmax (defined by Rowan-Robinson) is very sensitive to changes in the limiting magnitude assumed for the sample, and (ii) the redshift corrections (K-corrections) which have to be applied to the apparent magnitudes of the fainter radio galaxies (for which the evolutionary effects appear to be strongest) are not sufficiently well known. These effects will now be considered in turn.