This paper applies conditional quantile regression to specific production costs using the European Farm Accounting Data Network (FADN). In this process, specific production costs for each farm are assumed to depend upon the values of outputs broken down into sixteen categories of agricultural products. On the basis of results obtained for three major agricultural commodities produced in the European Union (EU) – namely wheat, cow milk and pigs – we are able to show the relevance of the quantile regression estimation procedure to handle the asymmetry and heterogeneity of the distribution of specific production costs in European agriculture. The conditional quantile regression was applied to specific production costs using 2006 FADN data for twelve EU member countries (Austria, Belgium, Denmark, France, Germany, Hungary, Italy, Netherlands, Poland, Spain, Sweden and United Kingdom). Econometric results show that significant differences among the twelve selected countries emerge for the estimated unit specific costs obtained for wheat, cow milk and pig.