Eight pre-concentration techniques were compared for their capacity to retain volatile and semi-volatile solutes during evaporation of solvent (dichloromethane). The 2-ml test-samples containing 0.2 ppm or 2 ppm (v/v) of volatile and semi-volatile solutes were concentrated to a final volume of 1 ml, 200 microl and 50 microl, respectively. When pre-concentrating to 50 microl, the highest recoveries for both the diluted (0.2 ppm) and concentrated (2 ppm) solutions were found by passive evaporation in a test tube at 22 degrees C. The pre-concentration time from 2 ml to 50 microl by this method was 19-20 h. Heating the test tube to 47 degrees C yielded lower recoveries in dilute samples, but the recoveries of concentrated samples were only slightly lower than the recoveries obtained by passive evaporation. The evaporation time was decreased to 1-2 h. The recoveries and the reproducibility of these methods were superior to the other pre-concentration methods tested. Loss of solute was apparently mainly caused by the fast vapour streams created when speeding up the process of evaporation by heating or by introducing a gas stream into the tube. This increased co-evaporation and thereby solute loss. The capacity of the methods to trap the escaping vapours and create a reflux determined the capacity of the methods to recover the solutes. The experiments demonstrated that more solute is lost during the pre-concentration of dilute samples compared to more concentrated solutions.