Thirty different eucalyptus oil samples were scanned on the FOSS NIRSystems 6500 Rapid Content Sampler using a reflectance vessel as sample presentation method. The cineole content of each sample was determined by the BP method and these reference data were used to construct two calibration equations for cineole content in the oils using Vision software. The mean accuracy for the NIR method differed by 1.01% or less, and the mean bias by +/-0.33% or less, compared with the BP method. Calculation of the 95% confidence intervals for the slope and intercept of plots of NIR predicted values against BP method reference values showed that there was no evidence of fixed or relative systematic errors. Tests for short-term and intermediate repeatability were conducted. The standard deviation was 0.83% w/w or less and the coefficient of variation was 1.11% or less. The confidence intervals for both short-term and intermediate repeatability overlapped with that for the BP method, suggesting that there was no evidence for a difference in values obtained by the BP and NIR methods. The range of cineole contents used in the calibrations was extended by incorporating five samples of eucalyptus oil spiked with cineole, and five samples of two essential oils known to have a lower cineole content than eucalyptus oil, to give a range of 52.5 to 99.0% w/w. The mean accuracy decreased to an error of 1.26% or less and the bias to +/-0.50% or less. Again, confidence intervals suggested there was no evidence for fixed or systematic errors in the NIR calibrations. We propose that NIR spectroscopy could be used as an alternative method for the determination of cineole content in eucalyptus oils.