INTRODUCTIONThis paper and its predecessor (Helliwell, 1996b) were triggered by the startling evidence, first presented by McCallum (1995), that 1988 merchandise trade flows among Canadian provinces were 20 times as dense as those between Canadian provinces and US states, after using a gravity model to control for the effects of size and distance. Such a striking finding demands verification and explanation. To see if similar results applied to other countries, Wei (1996) uses an imaginative method to generate approximate internal trade volumes and distances for OECD countries and finds for these countries a border effect that is statistically significant but far smaller than that found by McCallum. Head and Ries (1997) ask whether international migration may enable international trade opportunities to be better recognized and exploited. This paper is an attempt to build bridges between their research and the earlier findings. 1Abstract. The paper confirms a strikingly large effect of national borders on trade patterns. Estimates comparing trade among Canadian provinces with that between Canadian provinces and US states show interprovincial trade in 1988-90 to have been more than 20 times as dense as that between provinces and states, with some evidence of a downward trend since, owing to the post FTA growth in trade between Canada and the USA. Using approximate data for the volumes and distances of internal trade in OECD countries, the 1988 -92 border effect for unrelated OECD countries is estimated to exceed 12. Estimates from a census-based gravity model of interprovincial and international migration show a much higher border effect for migration than for trade, with interprovincial migration among the Anglophone provinces almost 100 times as dense as that from US states to Canadian provinces.