The trends and levels in childhood mortality from the cession of Fiji to the British in 1874 to the most recent census in 1986 are described and analysed. The core, and more detailed, analysis covers the period from 1893 -- when the first major Government inquiry into the decrease of the population was carried out -- to the late 1960s when relatively low mortality was reached. Both direct and indirect demographic methods are used to estimate levels of mortality. Explanatory material is drawn from a variety of documentary and ethnographic source describing the disease environment, public health programs, and changing technology. The social, cultural and biological determinants of infant and child mortality are considered. Special attention is given to the role of the Colonial Government, in public health policy, in the use of medical interventions, and in the establishment of a medical infrastructure. In the first decades of the 20th century the Government initiated a number of public health interventions and in the establishment of a medical infrastructure. In the first decades of the 20th century the Government initiated a number of public health interventions. The effects of these on child mortality are evaluated in the context of the disease environment. In the late 1920s and 1930s, they were supplemented by a child welfare campaign aimed at the active involvement of the village communities. Despite the modernity of this campaign its effects on child mortality were limited. The demographic analysis suggests that the "breakthrough" in Fijian child mortality, the period of steep decline over a relatively short time, came in the late 1950s. The factors contributing to this are explored and it is concluded that major elements in the accelerated decline were successive specific disease-related campaigns (particularly against tuberculosis and yaws) pushed through by an activist Government, making use of new medical technologies and drugs, and building on the decentralised medical infrastructure developed over the previous half-century. Fiji offers a rare opportunity for a study of this kind, since the Government was concerned from the outset of colonial rule with the problem of Fijian depopulation, and the place within it of child mortality; the need for an accurate system of vital registration was early recognised, and active programmes of public health and medical intervention pursued and well documented.