We provide recommendations for implementing telemetry studies on waterfowl on the basis of our experience in a tracking study conducted in three countries of sub-Saharan Africa. The aim of the study was to document movements by duck species identified as priority candidates for the potential spread of avian influenza. Our study design included both captive and field test components on four wild duck species (Garganey, Comb Duck, White-faced Duck and Fulvous Duck). We used our location data to evaluate marking success and determine when signal loss occurred. The captive study of eight ducks marked with non-working transmitters in a zoo in Montpellier, France, prior to fieldwork showed no evidence of adverse effects, and the harness design appeared to work well. The field study in Malawi, Nigeria and Mali started in 2007 on 2 February, 6 February and 14 February, and ended on 22 November 2007 (288 d), 20 January 2010 (1 079 d), and 3 November 2008 (628 d), respectively. The field study indicated that 38 of 47 (81%) of the platform transmitter terminals (PTTs) kept transmitting after initial deployment, and the transmitters provided 15 576 locations. Signal loss during the field study was attributed to three main causes: PTT loss, PTT failure and mortality (natural, human-caused and PTT-related). The PTT signal quality varied by geographic region, and interference caused signal loss in the Mediterranean Sea region. We recommend careful attention at the beginning of the study to determine the optimum timing of transmitter deployment and the number of transmitters to be deployed per species. These sample sizes should be calculated by taking into account region-specific causes of signal loss to ensure research objectives are met. These recommendations should be useful for researchers undertaking a satellite tracking program, especially when working in remote areas of Africa where logistics are difficult or with poorly-known species. ?? NISC (Pty) Ltd.