Much of the evidence underlying contemporary understanding and management of diabetes has accrued from clinical trials and observational studies which do not involve patient groups that are representative of those in the community, which have typically collected limited phenotypic data, and which are likely to involve follow-up which is short in relation to both the course of therapeutic progression and the development of potentially disabling and sometimes lethal complications. In light of these shortcomings, the Fremantle Diabetes Study (FDS) was started in 1993 with limited funding and basic infrastructure with the aim of identifying, and collecting detailed prospective data from, all known patients with diabetes living in a large, stable multi-ethnic urban Australian population around the port of Fremantle in order to facilitate an examination of clinically-relevant aspects including management, metabolic control, complications and cost. Over 20 years later, and with substantially improved finances and facilities, the FDS team have recruited over 3,000 representative patients in two phases (1993-1996 and 2008-2011) and developed an extensive database on each participant with longitudinal follow-up ranging from 3 to 21 years. Detailed analyses of the data, which includes linkage to local morbidity and mortality registers, have generated a wealth of information on the multi-system physical effects of diabetes, as well as its psychosocial and societal consequences. Original findings from the FDS and its sub-studies have had, and continue to have, important clinical and health-economic implications in Australia and beyond. The FDS illustrates what can be achieved with dedicated and skilled research staff and students, institutional and external support, and, most importantly, the enthusiastic involvement of the patients themselves who form the raison d’etre of the study.