The interest of the calibration of two chronological series of abundance index obtained from experimental and commercial fishing is useful because they are acquired in an independent manner one to each other, and in consequence, the sources of uncertainties and bias are different in these two experiences. In experimental fishing, the main source of mistake is linked to the small size of the sample. The sampled surface is very small, and moreover, the spatial heterogeneity of abundance index into the strata is important. In this case the imprecision of abundance index and their annual variability is higher than in the other case. In commercial fishing, fishermen tend to smooth the variations of abundance. The CPUE vary less quickly than the true variations of abundance. This bias is a general rule. Moreover, the yields are dependant of the variations of two factors : the catchability coefficient and the biomass. After having extracted of the yields the influence of the factors that have an effect on the catchability, the analysis shows that the common change of annual indicators of hake abundance reflects a stock change, in particular in the recruitment which depends on the environmental conditions.