For data-limited stocks (DLS), professional fishing data can be used as a potential source of information, especially in the absence of appropriate scientific survey, to understand abundance evolution under realistic hypothesis on resource catchability. This work focuses on a data-filtering approach, the selected fishing effort must reflect vessel activity that is least dependent on their technical characteristics and as stable as possible over time. The variable of interest is landing by fishing sequence (landing for a given gear, gear mesh, day and ICES statistical rectangle) called LPUE. In order to account for the abundance of the species, it was necessary to consider the discards. The method thus proceeds in 4 steps: (i) focus on LPUE variability and causes' prioritization; (ii) cluster definition to obtain a typology of vessels; (iii) average LPUE per cluster analysis; (iv) consideration of gear mesh classes and seasonal variations (quarters). This approach is outlined for the Striped red mullet of the Bay of Biscay that is currently in DLS category 5. Two reference fleets are thereby proposed: firstly the otter trawlers composed of small vessels (7.9–15.8 m) with a gauge of 2–43.9 grt, an engine power between 44 and 256 kW and a gear mesh of 70–79 mm; secondly the set gillnetters, which are defined by medium-sized vessels (8.2–14.8 m, 2–30.2 grt, 70–331 kW) whose gear mesh is either 50–59 mm (2nd and 3rd quarters), 60–69 mm (2nd quarter) or greater than 90 mm (2nd quarter). LPUEs of these fleets show a downward trend, significant in two out of four cases which may reflect a deterioration of the status of the Striped red mullet stock.