Thenium isotope ratio determinations are, in principle, possible by negative thermal ionization mass spectrometry (NTI-MS). Relatively high rhenium blanks from the commonly-used filament materials prevent accurate isotope ratio determinations, especially for small rhenium sample amounts which are of importance, for example, in geochronology in connection with the Re/Os dating method. Platinum and nickel filaments were tested by different preparation techniques to reduce the rhenium blank contribution from the filament material. The lowest rhenium blank of less than 1 pg was achieved by coating nickel filaments with V 2O 5 prior to degassing under high vacuum conditions at 850°C. Obviously, the vanadium—nickel oxide layer formed on the surface of the filament during this process prevents further emission of rhenium ions from the filament material. Using Ba(OH) 2 for the enhancement of negative thermal ions, 1 ng of rhenium resulted in ion currents at the detector side of about 10 −11 A with an ionization efficiency of up to 20%. The 185Re/ 187Re isotope ratio of a sample of natural isotopic composition could be determined to be 0.59818 ± 0.00026 with a relative precision of 0.04%. The isotope ratio determination for an 187Re spike was comparable in precision but the relative standard deviation of an 185Re spike was significantly higher, which could be explained by mass fractionations of oxygen in the measured ReO − 4 ion. The ReO − 4 ion is about 200 to 2500 times more abundant than the only other detectable rhenium ion in NTI-MSReO − 3. The ReO − 4/ReO − 4 ratio decreases with increasing temperature. By the low blank NTI-MS technique described in this work, more precise and accurate determinations of the rhenium isotope ratio and the rhenium concentration by isotope dilution analysis from nanogramme samples are possible.