This paper is a rejoinder by Professor Peter R. Odell to 'The future of oil' by H. R. Warman which was published in the September 1972 issue of this Journal. The paper presents an alternative interpretation of the nature and size of the world's proven oil reserves, their likely appreciation over time and the oil industry's ultimate resource base. It concludes that there is no evidence that there will be any physical shortage of oil to meet the world's needs over the next 50 to 60 years. The paper is followed by 'Further remarks on the future of oil' by R. D. H. Simpson which comments on Professor Odell's conclusions. IN his article, 'The future of oil' (Geogrl J. 138, 3), H. R. Warman justifiably observed, 'in our energy hungry society oil has assumed a dominant role with its future affecting economics, politics and the patterns of life' (p. 295). It is because oil has come to fulfil this role that, as with the question of war which is too important to be left to the generals, the future of oil is similarly too important to be left only to the oil companies. Hence this lengthy rejoinder on a subject which needs to be opened up to more general debate. During the past two decades of very high annual rates of increase in oil consump? tion in Western Europe, Japan and elsewhere, the spectre of the imminent danger of the exhaustion of conventional oil resources has often been raised?particularly by spokesmen for coal industry lobbies anxious to find arguments in favour of the pro? tection of indigenous but high cost coal (Odell, 1963). Until very recently, however, such arguments were constantly derided by the oil industry whose spokesmen were emphatic in their response that no restraint should be introduced on the growth of oil demand and who pointed out that the argument for limitation emerged from generally inadequate and incomplete evaluations of the nature of the oil resource base. In as far as the resource-exhaustion spectre was raised largely on the basis of a comparison of the current year's oil production with the so-called proven reserves of oil, the evaluations could certainly be described as inadequate and incomplete, for such proven reserves represent nothing more than the working stock of the oil indus? try. Thus, proven reserves can only be defined as a function of the demand for oil, given the fact that proving it is so expensive that no one?neither companies nor governments?has any incentive to undertake the task merely to show that the world, or a particular part of it, is oil rich. Oil is sought?and found?because oil companies and state entities have to make long-term investment and infrastructure plans to produce, transport, refine and market the oil they expect to sell and thus must have a firm idea as to where their crude oil supplies are to come from over the next ten years. Moreover, given their expectation, based on their experience over the period since 1945, that the demand for oil will build up year by year they know that the reserves available to them on which to justify such long-term investment must be equal to about 15 times the current annual level of consumption. This, in theory, sets what might be termed the overall optimum level for proven reserves. In practice, in a world of less-than-perfect mobility for the trading of commodities and divided into hostile or potentially hostile groups, the search for security of oil supply will necessitate the discounting of the apparent availability of some resources and the discovery of others in areas on which greater strategic and political reliance can be placed. This is one of the two main reasons why the optimal reserves/production (R/P) ratio has been greatly exceeded throughout the post-war period (Fig. 1). It should be noted that it was not generally exceeded before 1940 when the proven reserves of the ^ Professor Peter R. Odell is Director of the Economisch-Geografisch Instituut, Nether? lands School of Economics, Rotterdam,