European legislation has set a limit of 5 mg Hg per compact fluorescent lamp (CFL). Compliance with this regulation is tested with a method given in a European law from 2002 ("EU method"). According to the EU method, the arc tube has to be placed inside a fume cupboard and cut into segments. These pieces are collected and washed with HNO3, which is finally analyzed for Hg. In this study, we investigated the losses of Hg via the gas phase, which are ignored by the EU method, using a semiquantitative radiochemical neutron activation analysis (RNAA) technique developed at the TRIGA Mark II reactor in Vienna (Austria). Depending on the product, the losses range as high as approximately 2% at 20 °C. At higher ambient temperatures (hot summer days), the losses may increase by a factor of 4. Quality products generally suffer lower emission via the gas phase than cheap products, which is probably due to amalgam or other techniques that retain the Hg from broken CFLs. In any case, the EU method introduces a small but avoidable systematic error into the analysis and, in contrast to recommended practices, underestimates the Hg content of the CFL in question. Having technical alternatives at hand, we urge European lawmakers to adjust the law accordingly.