During the 1928-29 Expedition, centred at Low Isles, Spender mapped the 'low wooded islands' or 'island-reefs' of Low Isles and Three Isles in detail, and additional information was published by Steers, T. A. Stephenson and others. From this work, two different models of the evolution of low wooded islands were proposed, Spender holding that the islands were in a state of equilibrium resulting from their location on the reef, Steers that they could be placed in an evolutionary sequence. Moorhouse described the results of cyclones at Low Isles in 1931 and 1934, and Fairbridge & Teichert reconsidered the general issues following aerial reconnaissance and a brief visit to Low Isles in 1945. Subsequently, aspects of change since 1928-29 have been studied at Low Isles by W. Stephenson, Endean & Bennett in 1954 and by W. Macnae in 1965. Maps produced since 1929, however, have all been based on Spender's surveys. In 1973, Low Isles and Three Isles were remapped in detail, and a direct comparison can now be made over an interval of 45 years. This shows changes in island topography, and substantial alteration in the size and location of shingle ramparts which has affected conditions for coral growth on reef flats. Mangroves have extended greatly at Low Isles, but not at all at Three Isles. The implications of these findings for the general models of Steers and Spender will be discussed and related to the Holocene history of the Great Barrier Reefs.