An Early Proterozoic Kikkertavak mafic dyke intruding the Archaean Hopedale block, Labrador, gives an age of 2235 ± 2 Ma using U–Pb techniques on baddeleyite. A Harp mafic dyke in the same area gives an age of 1273 ± 1 Ma using U–Pb techniques on baddeleyite and zircon. The latter age is almost identical to that of the giant Mackenzie swarm and to the age of the BD0 dykes in South Greenland, and points to a major pulse of mafic magmatism over much of the North Atlantic craton at this time. The former age is a little older than available Rb–Sr ages for the extensive MD swarm in West Greenland, but there are possible correlatives.Geochemical data are presented to ascertain whether there are significant compositional differences between the Harp and Kikkertavak dyke swarms. In fact, two distinct chemical subgroups can be recognized within the Kikkertavak dykes, and three others are recognized within the Harp suite. These differences apply more to trace element patterns rather than major element abundances, but although there are compositional differences between the average Harp and average Kikkertavak dyke, it is unlikely that geochemistry could be used unequivocally to separate the two. The compositional differences probably reflect evolutionary processes in the lithosphere. The range of composition exemplified by the subgroups is most easily interpreted in terms of proportion of asthenosphere and lithosphere components, and does not necessarily imply that either dyke swarm was emplaced over a long period. The presence of subgroups within both swarms urges some caution in assuming all dykes correspond to one or other age.