From 1962 to 2002, at the Saraoutou research station on the island of Santo in Vanuatu, genetic improvement work was undertaken to increase the productivity of coconut plantations, which, through copra exports, were the country's main wealth in the last century. Beginning with two Vanuatu Tall (VTT) populations collected near the station, four mass selection cycles by open pollination or intercrossing have resulted in Elite Vanuatu Tall populations intended for distribution to farmers. The study sums up how the different VTT populations perform in terms of germination rate, flowering precocity, and yield components (copra per nut, number of nuts). It highlights the efficiency of these breeding methods in increasing nut copra content and in reducing within-population variance of phenotypic traits. Flowering precocity, the number of nuts, and consequently copra production per plot, remain highly dependent upon growing conditions, and on the care taken in the nursery. It also confirms that selection too exclusively based on the search for a high copra content is reflected in a drop in the number of nuts and a very slight gain in copra production per palm. Compared to higher-yielding hybrids, the improved VTT populations offer the advantage of being totally tolerant of coconut foliar decay and of being reproducible by farmers themselves. The merits of setting up decentralized seed gardens in the Vanuatu archipelago from improved populations at the research station, or from locally surveyed material, are discussed.