The general aim of the ECOKNOWS project is to improve the use of biological knowledge in fisheries science and management. The lack of appropriate calculus methods and fear of statistical overparameterisation has limited biological reality in fisheries models. This reduces biological credibility perceived by many stakeholders. We solve this technical estimation problem by using up-to date methodology supporting more effective use of data. The models suggested will include important knowledge about biological processes and the applied statistical inference methods allow to integrate and update this knowledge in stock assessment. We will use the basic biological data (such as growth, maturity, fecundity, maximum age and recruitment data sets) to estimate general probabilistic dependencies in fish stock assessments. In particular, we will seek to improve the use of large existing biological and environmental databases, published papers and survey data sets provided by EU data collection regulations and stored by ICES and EU member countries. Bayesian inference will form the methodological backbone of the project and will enable realistic estimations of uncertainty. We develop a computational learning approach that builds on the extensive information present in FishBase (www.fishbase.org).The developed methodology will be of fundamental importance, especially for the implementation of the Ecosystem Approach to Fisheries Management. It has been a difficult challenge even for target species with long data series, and now the same challenge is given for new and poorly studied species. We will improve ways to find generic and understandable biological reference points, such as the required number of spawning times per fish, which also supports the management needs in the developing countries. ECOKNOWS applies decision analysis and bioeconomic methods to evaluate the validity and utility of improved information, helping to plan efficient EU data collection.