The altitudes of dated raised shorelines in parts of Fennoscandia result from the algebraic sum of isostatic emergence and eustatic submergence. As the isostatic emergence differs systematically from place to place, the eustatic rise can be calculated within narrow limits. Data from New Zealand, some of them new, together with the calculated sea levels with which they are in general agreement, support Fairbridge's recent deduction of a Postglacial sea level higher than the present. In detail, however, Fairbridge's curve of sea-level rise seems to require revision, notably in the elimination of a −15 metre level 9,000 years ago and-possibly also of a +3 to +4 metre level, 5,000 years ago. Acceptance of the calculated sea levels permits computation of the real uplift for localities on the Baltic and Canadian shields. When uplift is plotted against time, the curves obtained for all localities belong to one family, indicating that similar forces were at work at precisely the same times within both shields. This similarity, together with the high rate of uplift, particularly in early Postglacial time, makes it almost certain that isostatic rebound, following melting of the ice, has been real and not illusory.