Based on some round robin experiments organized in the last years by a DIN working group and by the IUPAC working party IV.2.2, the paper discusses measures for getting a better interlaboratory reproducibility of GPC results. The most important points are: selecting columns according to their separation selectivity and range avoiding artifacts in the distribution curves, caused by gaps in the pore volume distribution avoiding adsorption and memory effects calibration: using calibration standards of high quality selecting suitable calibration points checking the influence of the injected concentration selecting a common detector data treatment, especially finding the correct baseline preparing a detailed data report according to a uniform schedule.
