The fatty acid (FA) content of fish is generally said to reflect fatty acid composition of the diet. In fact, incorporation of FA into tissues is modulated by various metabolic factors, and final composition will depend upon the initial FA content, cumulative intake of dietary fatty acids, growth rate and duration. Analysis of time course of changes in FA composition should be easier with animals having different initial FA profiles, which are subsequently fed a diet with the same FA composition. Data from two studies, one with brown trout and another with turbot were used. Fish were first fed with diets containing one of three different oils (soybean oil (SO), linseed oil (LO) and fish oil (FO)), and subsequently fed the same fish oil-based diet (washout period). If we suppose a model fish having the same initial composition as those fed vegetable oil and which incorporate fatty acid in the same way as the control fish always fed fish oil, we may compute a model of dilution of initial fatty acid content with increasing growth and absolute fat deposition. Experimental data can be compared with a reference fatty acid profile given by this model for the same fatty acid increase. Application of the model to experimental data shows that while muscle neutral lipid (NL) FAs roughly follow this dilution model, those of muscle polar lipids (PL) undergo much faster changes than model values based on increase of total polar lipid quantities. Among observed differences between the model and experimental values, DHA is of particular interest as this fatty acid displays lower change rates (significant in turbot neutral lipids) than expected in contrast to other fatty acids.