Although a tendency for high latitude marine invertebrates to avoid pelagic larval stages was first described in the 19th century, the most detailed early study was that of Thorson in Greenland. This work also established other features of the reproduction of polar marine invertebrates that have become regarded as almost axiomatic (e.g., the release of larvae to coincide with the summer bloom) or largely ignored (a latitudinal cline in egg size within species). This short and selective review examines Thorson's conclusions in the light of recent work. It is now clear that although polar prosobranch gastropods reproduce almost entirely by direct development, for many taxa the real distinction between polar and non-polar species is in the proportion of feeding to non-feeding larvae. Some species release feeding larvae in winter and the energy source for these larvae is obscure. Growth is slow and there is little or no evidence for temperature compensation. Many crustacean species have larger eggs at higher latitudes. Egg size varies significantly within species, with larger eggs being associated with larger females and often reduced fecundity. The reasons for these within-species patterns are currently unresolved.