An objective of the Atlantic Ocean Tropical tuna Tagging Programme (AOTTP) was to estimate Type-I (immediate) and Type-II (long-term) tag-shedding rates for tropical Atlantic tunas from double-tagging experiments. Historical information on tuna tag-shedding studies conducted in different parts of the world was incorporated as prior distributions using a Bayesian approach to estimate the new tag-shedding parameters. Type-I and Type-II tag-shedding rates were respectively estimated at 0.007 and 0.084/yr for bigeye tuna, 0.021 and 0.051/yr for skipjack and 0.021 and 0.088/yr for yellowfin tuna. Using realizations derived from the MCMC posterior distributions, the shedding rate was estimated to reach 50% of the tags after seven and a half years at sea for yellowfin and after eight years at sea for bigeye tuna. The loss rate of conventional tags is lower for skipjack. Our results suggested that continuous Type-II shedding rate is size-dependant for yellowfin and bigeye (i.e., showing a three-fold increase between individuals less than 45 cm fork length (FL) at release and fishes larger than 65 cm FL). This study reinforces the need to account for tag-shedding along with other sources of uncertainty, such as reporting rate, in order to accurately estimate the exploitation and mortality rates derived from tagging data.