White spot disease (WSD) is caused by white spot syndrome virus (WSSV) and is an acutely fatal pandemic disease of crustaceans. It has resulted in massive losses to the shrimp-farming industry in Asia and has now spread to the Americas. This paper reports the problems and solutions associated with the design and execution of a longitudinal epidemiological study of shrimp (Penaeus monodon) health on farms practising a crop rotation of rice and shrimp in the Mekong Delta of Vietnam. The pre-sampling phase of the project involved selecting an appropriate site and sampling variables, obtaining permission and establishing the necessary laboratory and logistic facilities. At the start of the sampling phase, 40 farmers were selected and 32 of these were visited and interviewed. This resulted in the enrolment of only 17 farmers. A further seven had to be enrolled to obtain the maximum number of farmers that could be sampled by the study team. Compliance was enhanced through meetings, regular visits by senior members of the project team and ensuring that visits were punctual and that all information was treated confidentially. The production cycle began in January 1998 and lasted for approximately 5 months. An attempt was made to collect 500 post larvae (PL) before each pond was stocked to assess the health of the batch and to test for the presence of WSSV by one-step PCR. After stocking, the wild crustaceans also were sampled from the pond for PCR analyses. Information was collected on the management practices and samples of water, pond bottom, feed and shrimp collected throughout the production cycle. Water quality variables with predictable diurnal variation were sampled in the morning and afternoon, twice a week. Two months after stocking, the first outbreak of WSD occurred; subsequently, 18 farms conducted a complete emergency harvest due to the actual or perceived presence of a WSD outbreak. Detectable mortalities were reported from 19 farms, and moribund shrimps were collected from four of these for PCR and histological analyses.