Ten years after the start of the European Data Collection Framework the availability of cost indicators for assessing the economic status of fisheries or bio-economic modelling is still deficient. Moreover, economic time series are difficult to maintain due to fishermen weariness and sample coverage is often insufficient. To overcome these problems, the paper builds predictive models per operational cost category (fuel, landing, other variable, fixed, salary and total) for the French Bay of Biscay fleet. The cost models had good fit and stressed the key role of gear and vessel size (length mostly) for explaining fishing costs. They suggested that only a small list of variables was needed to predict costs (days-at-sea, total revenue, vessel age, length and power, district and fishing zone). These variables are easy to obtain for all vessels. In addition, some fitted cost models (total, landing and fixed costs) could be applied to the neighbouring English Channel, which suggested similarities between their fishing cost structures. Finally, the comparison of design-based and model-based cost estimates for the Bay of Biscay French fleet confirmed that modelling fishing costs on the basis of a small set of explanatory variables is a good alternative to design-based estimates which generally require costly annual sampling programs.