Two different mass spectrometric methods, inductively coupled plasma isotope dilution mass spectrometry (ICP-IDMS) and negative thermal ionisation isotope dilution mass spectrometry (NTI-IDMS), were applied to the determination of trace concentrations of boron in rainwater samples. After carrying out the isotope dilution step, ICP-MS was directly used for the measurement of the ¹¹ B/ ¹⁰ B ratio. For NTI-IDMS, the isotope diluted boron was first separated by a boron selective ion exchanger and then the boron isotope ratio was determined by measuring negative BO ₂ ^– thermal ions. Boron concentrations of 21 rainwater samples from different parts of Germany and from the Jungfraujoch in Switzerland at an altitude of about 3600 m were determined to be in the range 0.3-7 ng ml ^–1 . The detection limits were 0.2 and 0.3 ng ml ^–1 for ICP-IDMS and NTI-IDMS using 3 and 50 ml sample volumes, respectively. NTI-IDMS determinations were about a factor of 2.5 more time-consuming compared with ICP-IDMS. However, the boron concentrations determined by ICP-IDMS often showed slightly higher values than those determined by NTI-IDMS, which was due to contributions of relatively high ¹² C ⁺ ion intensities to the ¹¹ B ⁺ intensity when using a quadrupole ICP-MS instrument with insufficient mass resolution. ICP-IDMS is a fast and reliable method for the determination of boron concentrations down to <1 ng ml ^–1 in aquatic systems if mass resolution between mass numbers 11 and 12 is under control, whereas NTI-IDMS can be best applied as an accurate reference method at these low concentration levels. Significant dependence of the boron concentration in rainwater on the meteorological conditions, but not on the season of sampling, as well as anthropogenic influences, especially on the altitude of the sample collection place, was found by applying both mass spectrometric methods.