An accurate understanding of the leeway drift characteristics of drifting objects is required to effectively forecast the drift of persons, vessels or objects lost at sea, and to generate efficient search areas to maximise the probability of successfully locating those missing. Presently, the most effective method for calculating the leeway drift characteristics of an object or vessel is to empirically derive the leeway coefficients of that object through field studies. The main goal of the studies is to measure how the object drifts in relation to the surface currents, due to the wind and wave action upon it. This paper outlines the determination of downwind and crosswind leeway coefficients for three small craft common to Pacific island communities for which no accurate leeway coefficients exist. These craft were: a 19 foot (5.8m) fibreglass skiff (known locally as pangas, fibres, or banana boats); a 20 foot (5.97m) fibreglass outrigger canoe; and a 2-person sit down personal water craft (PWC). Due to the vast distances between pacific islands and the remoteness of these locations it can be several days until a search can be mounted to rescue those lost at sea, hence it is paramount that an accurate description of the drift of these tropical pacific craft is available for use in search and rescue (SAR) drift models, to define appropriate search areas. This study successfully derived the leeway coefficients required for each of these three craft. The leeway speed of the outrigger canoe and PWC, both with one person on board (POB) equivalent loading, were calculated to be 2.40% and 4.24% of the wind speed respectively. The leeway speed of the skiff was found to range between 7.71% and 4.40% of the wind speed for equivalent loading between 1 POB and 13 POB.The results of these field tests have subsequently been implemented into search and rescue models by several SAR organisations worldwide. These results show that the findings herein have the potential to both increase the likelihood of finding persons adrift at sea alive, as well as reducing search costs through more effective drift prediction and efficient search area formulation.