Three different extraction methods for Turbellaria were compared, 2 of which use living material. The first method is based on sea-water ice treatment (Uhlig-method), the second employs decantation of narcotized fauna (MgCl2-method). The third method is the Barnett-method, performed on preserved material. Three series of 12 samples each were analysed quantitatively. Turbellaria were counted and identified to the order level. For the turbellarian population as a whole, the numbers of individuals extracted alive (Uhlig- or MgCl2-method) was less than 50% of the number extracted from the preserved samples (Barnett-method). Differences between Uhlig-method and MgCl2-method were not statistically significant. However, for Acoela and Retronectida, the Uhlig-method was more efficient than both other methods. We conclude that for a reliable quantitative analysis of Turbellaria 2 different extraction methods are required: one should employ extraction of living material (Uhlig-method) and one extraction of fixed material. For qualitative investigations extraction of living material is recommended.