THE ISLAND of Barbados has been fully cultivated since the mid-seventeenth century and has been considered over-populated for three hundred years. Sugar-cane was introduced in I640 and has remained the mainstay of the economy of the island ever since providing, in I963, two-thirds of the exports by value. Today, Barbados, an island the size of the Isle of Wight, has a population of nearly one-quarter of a million at a density of 1458 per square mile or 3.6 persons per acre of cultivated land. In 1945, when the population density was only 1163 per square mile, E. W. Gilbert and R. W. Steel stressed the need for a geographical study of population in this island with 'the greatest rural density in Central and South America'.1 Twenty years later population density is still increasing and the safety valve of migration has been almost shut off by recent British immigration policy. In the face of falling sugar prices the government and planters of Barbados have finally accepted the need for a radical reconsideration of the basis of the island's economy. Tourism and industry are being developed but agriculture remains important, the main emphasis being placed on diversification of crops and the contribution of peasant farmers to local food production. Consequently it is particularly relevant at this time to look at the inter-relationships of population and agriculture in Barbados. Two hypotheses are advanced and tested. First, it is suggested that population variables account for a significant proportion of the internal variation in the structure of peasant farming in Barbados. A multivariate statistical technique, factor analysis, is used to test this. Secondly a hypothesis put forward by T. Lynn Smith is examined in the Barbadian context.2 He states that in areas dominated by a plantation economy the dynamic aspects of population have certain characteristics, and, in particular, where 'large sugar plantations dominate the scene the age profile takes on the characteristic contours of an urban area'.3 It was felt that it would be valuable to test this hypothesis in Barbados where, according to the I96I Agricultural Census, 82 per cent of the farmland was in sugar estates of more than a hundred acres.