Using a newly developed technique, methylmercury species have been quantified in several precipitation and lake water samples. Mercury species are determined by aqueous-phase ethylation to the volatile dialkyl analogs, followed by cryogenic gas chromatographic (GC) separation. Mercury-specific detection by cold vapour atomic fluorescence spectrometry permits detection limits of about 0.1 pg as Hg. Snow samples collected from north-central Wisconsin contained monomethylmercury levels of about 0.25 pM (0.05 ng l?1 Hg) and total mercury concentrations of 20 pM (4 ng l?1 Hg). A time series of rain samples collected during a storm passing over the North Olympic Peninsula in western Washington State showed average monomethylmercury levels of 0.75 pM (0.15 ng l?1 Hg), with total mercury concentrations from 10 to 25 pM (2?5 ng l?1 Hg). Total mercury showed a strong washout effect over the course of the storm, while methylmercury appeared to show a diurnal pattern, with elevated levels during the daylight hours. No dimethylmercury was observed in any precipitation sample. Methylmercury was observed in most lakes studied, with a high of 3.1 pM (0.64 ng l?1 Hg) in Onadaga Lake, New York, and a low of