Re-evaluation of the age of Canadian dike swarms using Rb–Sr whole-rock data give the following results:Detailed studies of the Matachewan dike swarm show that in most cases Rb–Sr mineral isochron age values from individual samples are concordant with the Rb–Sr whole-rock age values for the dike swarm as a whole and are clearly greater than K–Ar age values on the same material. The fact that the mineral isochron ages and the whole-rock ages by the Rb–Sr method are similar suggests that there has been no significant thermal overprinting of these particular dikes since their emplacement and solidification or contamination by radiogenic 87Sr from the host environment. Furthermore, the general scatter and lack of significant grouping of the lower K–Ar values similarly suggests that there was no single period of thermal resetting of the K–Ar ratios. It is observed that the K–Ar age values are greater in samples containing mica, and are more scattered and lower in samples in which potassium is contained in late-stage or principal phases other than mica. Therefore, it seems likely that significant loss of argon may result at near or normal surface temperatures and that the rate of diffusion is dependent on the mineralogy of the sample. It is concluded that reasonably good age determinations can be made on diabase dike rocks of ancient age by the Rb–Sr whole-rock isochron method, if the mineral isochrons are concordant.