This is the first publication which describes the evaluation of the analytical performance and state-of-the-art of the determination of methyl tert-butyl ether (MTBE) in water at ng L(-1) concentrations. An interlaboratory comparison study for the determination of MTBE in water was carried out. Twenty-eight laboratories from seven European countries participated in the study. Twenty of those finally transmitted results to the organiser. Italian spring water, containing no detectable amounts of MTBE was fortified to yield two samples with MTBE concentrations of 0.074 +/- 0.004 microg L(-1) and 0.256 +/- 0.010 microg L(-1). The laboratories applied their regular in-house methods to analyse the water samples. Static headspace, Purge & Trap, solid-phase microextraction (SPME) or direct aqueous injection were used as sample preparation techniques. Subsequent separation and detection of MTBE were performed by gas chromatography/mass spectrometry (GC/MS) or gas chromatography/flame ionisation detection (GC/FID). After rejection of outliers, the overall arithmetic mean of laboratory results corresponded to recoveries of 78 +/- 20% (Sample A) and 88 +/- 20% (Sample B) of the reference concentrations. The between laboratory coefficients of variation (CV) were 32% and 31%, respectively. The organisation of the study and quality assurance measures at the organiser's laboratory are described. Moreover, the measurement results of the participants and the analytical methods used for the determination of MTBE are presented and the correlation between selected method parameters and data quality is discussed.