Since the beginning of PV module production, Soxhlet extraction has been the standard method for the determination of the gel content and the resulting calculation of the degree of crosslinking of the most common PV encapsulant ethylene vinyl acetate (EVA). While the method is deemed well established and several pertinent ASTM and ISO standards exist, in practise the actual procedures used in - and trusted by - both industry and R&D institutions vary substantially. To evaluate the reliability of the methods and the comparability of the results, a round-robin test involving seven independent European laboratories one industrial PV module manufacturer and six R&D facilities - was conducted. The measurements were performed using homogenous, anonymized sample sets, each comprising five differently crosslinked EVA foils. The analysis showed that results obtained for the same samples may deviate significantly, but also that very different analytical procedures can yield comparable values. In a systematic study, the impact of various key parameters of the analytical process (extraction time and solvent, drying conditions, sample size and weight etc.) was investigated. Based on these findings, deviations observed in the round-robin study could be linked to their origins and the main pitfalls were identified. In conclusion, a suggestion for an optimised standard procedure was derived to ensure comparable results at all laboratories. (C) 2015 Elsevier B.V. All rights reserved.