Tropical soil science was founded on soil survey. Identification and mapping of the main soil types of a country was the first step towards evaluation of their land use potential, leading to agricultural development. With the major exceptions of India and Ceylon (Sri Lanka), most early soil survey was conducted by expatriate staff from the colonial powers. The leading players were Dutch, French, Belgian and British scientists. The Dutch in the East Indies (Indonesia) produced the first tropical soil map in 1901, for part of Sumatra. French studies in West and Central Africa were conducted by ORSTOM (founded, as ORSC, 1943), with Georges Aubert a leading figure. Belgian contributions were the founding in 1953 of the Inter-African Pedological Service of the CCTA in Yangambi, Belgian Congo (Congo), a project of which was the first Soil Map of Africa (1964) by J. L. d'Hoore. There was a US survey of Puerto Rico in 1942, and collaboration of American scientists with South America and with China. Further links were between the former USSR and developing countries of Central Asia. Soil survey in the former British colonies falls into three eras: pioneers, the age of reconnaissance surveys, and post-independence studies by local staff. The pioneers, who began surveys between 1920 and 1932, were Frederick Hardy in the West Indies, Arthur Hornby in Nyasaland (Malawi), Frederick Martin in Sierra Leone, Geoffrey Milne in East Africa, the ecologist Colin Trapnell in Northern Rhodesia (Zambia), and Cecil Charter initially in the West Indies but most notably in the Gold Coast (Ghana). These are giants on whose shoulders their successors stand. The era of reconnaissance surveys was 1950-1970. Incentives were the post-war focus on development, the UK Colonial Development and Welfare Act (1945) and Colonial Pool of Soil Surveyors (1954). Rapid survey became possible through availability of air photographs, and application of a geographical or ecological approach. Space restricts mention to only a few countries. The most complete national coverage was of Uganda, carried out largely 1958-60. Two of the three provinces of Nyasaland (Malawi) were surveyed (by the writer) 1958-62. From 1951 onwards there were surveys of large parts of the Gold Coast (Ghana) and Nigeria. Immense surveys, by grids of pits, were conducted for irrigation schemes in Sudan. The UK Land Resources Development Centre surveyed not only soils but also forest and pasture resources, extending resource surveys to land evaluation and, in a few cases, economic analysis of development options. Work in the West Indies was based on a single Regional Research Centre, which produced 26 ‘Green Book' surveys. National cross-fertilization came through New Zealand cooperation (1964-67) with Malaya (Malaysia) and Pacific island territories, and Dutch aid to the Soil Survey of Kenya from the 1970s. By 1970 most countries had sufficient information to supply national soil maps, of varying reliability, to FAO-Unesco Soil Map of the World (1970-80). Small-scale maps were a starting point for national planning, and importantly, identification of soil types for subsequent more detailed surveys. The third era follows political independence and the replacement of expatriates by local professional staff. India is an exception in that staffing has always been largely local, with S. P. Raychaudhuri a leading figure; an All-India Soil Survey Scheme, a truly ambitious project, was started in 1956. Most countries retain a soil survey unit with staff of high quality, although their activities are often constrained by a low operating budget. Efforts are focused on local surveys for specific development purposes. Among early staff of local origin were Henry Obeng and Victor Adu in Ghana, and Harry Obihara in Nigeria. For priority, however, none can compete with Mohammed el A'al, appointed in Sudan in 1929, and A. W. R. Joachim, Ceylon (Sri Lanka) 1930. Presentation will focus on the era of reconnaissance surveys. Discussion is invited particularly on the aims and achievements of this period, and whether the results are of lasting value at the present day.