Survey indices play an important role in stock assessments as they provide information on stock trends. In certain cases large interannual variations have been observed which are unlikely to reflect true underlying stock changes but are rather outliers. When survey indices for several species appear to be outliers for the same year, the suspicion is raised that something happened during the survey of that year. This is called a year-effect in survey catches. To study the potential year-effect in survey catches for the French autumn groundfish survey taking place in the Bay of Biscay, several indicators for survey design and wind conditions were derived as explanatory variables and Principal Component Analysis was used to study the relationship between these variables. Using multiple linear regression models we found that, on average 20% of interannual variation in abundance indices could be explained by survey conditions for benthic species, 11% for demersal and none for pelagic species. In contrast, survey conditions explained a smaller and decreasing part of the interannual variability in the coefficients of variations of these abundance indices and in species mean weight for benthic, demersal and pelagic species. Thus survey indices of benthic species seemed most affected by survey design and wind conditions.