Fishing pressure on reef fish stocks, in relation to exploitable biomass, is a useful parameter for fishery resource management but difficult to assess in subsistence fisheries. A methodology for indirect fishery output assessment in island environments is proposed based on estimates of consumption and the relevant seafood trade. Ouvea atoll in the South Pacific (Loyalty Islands, New Caledonia), where subsistence fishing is the main activity, was used as the experimental site. A consumer survey of the islands' households was conducted in June 2000. The sampling plan was determined on the basis of data already available on the population concerned, making it possible to give quantitative and qualitative information on catches. Total catch is estimated to be 212 t?y?1 composed mainly of Lethrinidae, Serranidae and Gerreidae. Subsistence activity is the main source of supply for two-thirds of households. These results were validated by a survey carried out on fishers at the same time that confirms the reliability of fish consumption as an indicator of reef fish catches. This study also showed that shoreline fishing remains the main activity in Ouvea: it's practised by 71% of households and provides almost 48% of total landings. A second stage was the identification of variables liable to influence consumption patterns: fishing practice (frequency and gear type), income, and household size seemed to be the main determining factors. These variables were integrated into a multi-dimensional analysis in order to define types of consumers. This consumption-based approach underlies a sampling strategy that could be extrapolated to other small South Pacific islands. It involves advantages of various kinds ? accuracy, reliability, and cost-effective implementation ? that make it a powerful tool for monitoring island fishery development.