Despite a rapidly growing literature on analytical methods and field applications of O isotope-ratio measurements of NO3− in environmental studies, there is evidence that the reported data may not be comparable because reference materials with widely varying δ18O values have not been readily available. To address this problem, we prepared large quantities of two nitrate salts with contrasting O isotopic compositions for distribution as reference materials for O isotope-ratio measurements: USGS34 (KNO3) with low δ18O and USGS35 (NaNO3) with high δ18O and ‘mass-independent’ δ17O. The procedure used to produce USGS34 involved equilibration of HNO3 with 18O-depleted meteoric water. Nitric acid equilibration is proposed as a simple method for producing laboratory NO3− reference materials with a range of δ18O values and normal (mass-dependent) 18O:17O:16O variation. Preliminary data indicate that the equilibrium O isotope-fractionation factor (α) between [NO3−] and H2O decreases with increasing temperature from 1.0215 at 22°C to 1.0131 at 100°C. USGS35 was purified from the nitrate ore deposits of the Atacama Desert in Chile and has a high 17O:18O ratio owing to its atmospheric origin. These new reference materials, combined with previously distributed NO3− isotopic reference materials IAEA-N3 (=IAEA-NO-3) and USGS32, can be used to calibrate local laboratory reference materials for determining offset values, scale factors, and mass-independent effects on N and O isotope-ratio measurements in a wide variety of environmental NO3− samples. Preliminary analyses yield the following results (normalized with respect to VSMOW and SLAP, with reproducibilities of ±0.2–0.3‰, 1σ): IAEA-N3 has δ18O = +25.6‰ and δ17O = +13.2‰; USGS32 has δ18O = +25.7‰; USGS34 has δ18O = −27.9‰ and δ17O = −14.8‰; and USGS35 has δ18O = +57.5‰ and δ17O = +51.5‰. Published in 2003 by John Wiley & Sons, Ltd.