The work carried out in this thesis aimed to improve the understanding of effects of season and learning capabilities of fish on trawl cod-end selectivity. To investigate the seasonal variation in cod-end selectivity, three sea trips were carried out on board a Scottish commercial trawler. Experiments were performed on the same grounds using the same fishing gear with an improved covered cod-end technique. The target species was haddock (Melanogrammus aeglefinus). Sea trips were conducted in April 1995 (post-spawning stage), September 1995 (after summer feeding) and February 1996 (pre-spawning stage). Selectivity parameters were estimated haul by haul for each season by a generalised linear model with logistic link. The 50% retention lengths (L₅₀) for each cruise were found to be significantly different. The selectivity was poorest in post-spawning stage in April (L₅₀=27.7cm), the highest after summer feeding in September (L₅₀=33.4cm) and intermediate at pre-spawning stage in February (L₅₀=31.2). There were differences in the length-girth relationship of the haddock between the three cruises. However, the variation in selectivity between seasons cannot be explained by the variation in girth alone. The lowest L₅₀ occurred when girths at a give length were smallest in April. Girth measurement were obtained in September 1995 and February 1996 allowing the selectivity parameters to be estimated in terms of girth. The 50% retention girth (G₅₀) in September (G₅₀=177mm) was significantly higher than that in February (G₅₀=159mm). One of the potential reasons for this variation is expected to be due to the effect of the water temperature on the maximum swimming performance of fish. Laboratory experiments to predict the maximum swimming speeds of haddock and whiting (Merlangius merlangus) from the minimum muscle contraction time of the fish, when stimulated by an electric pulse, showed a significant increase of the maximum swimming speed at 12 °C in comparison to 7 °C.