The tagging data base compiled since the 1960s by the secretariat of ICCAT was used to assess the effects of several release factors on the recovery rate of bigeye tuna (Thunnus obesus). The release factors considered were the length of the fish, the latitude, the longitude, the gear used during the tagging operation, the kind of tag used, the year and the month as well as the interactions between the length and the tag type and the length and the year. It was showed that all of these explanatory variables influenced significantly the probability of return of the fish and were used within a GLM framework to predict the probability of recapture of the fish. Results from this model reinforce previous findings concerning the lower efficiency of Betyp tag compared with Spaghetti tag but also highlighted the fact that this detrimental effect concerns only fish lower than 80 cm (FL). Other aspects related to the presence of interactions and the difficulty to discern from inspection of the coefficients how some factors combine to influence the probability of recapture are also discussed.