Biomonitoring of mercury (Hg) in urine can provide valuable information on environmental exposure to inorganic and elemental Hg. However, this requires a method with a sufficiently low detection limit, and in addition, sample stability during storage needs to be addressed. We adapted a method described in the literature to obtain lower detection limits by decreasing dilution and optimizing the amount of reagents used, while also investigating matrix effects, within- and between measurement variability, and accuracy. We tested Hg stability in urine under different storage conditions and using different stabilizers suggested in the literature. A five-fold dilution of the urine sample with addition of optimized amounts of BrCl gave the best results regarding detection limits, which could be further improved by using a gold amalgamation preconcentration step (Autrap). Application of the method in a biomonitoring study allowed detection of Hg in 95% of the samples without Au trap, and resulted in mercury levels that were comparable to similar populations and influenced by expected variables (age, dental amalgam). Regardless of the sample storage conditions of urine samples in polypropylene bottles, Hg decreased during the first days of storage, then slowly regained or exceeded the originally measured concentration. From these experiments, it appears that storage of the samples frozen at -20°C for several weeks prior to analysis, without the addition of additives is the preferred method.