The identification of Melaleuca cajuputi leaf samples (trees) that demonstrate enhanced oil characteristics using near infrared (NIR) spectroscopy is described. Leaf samples from an unthinned M. cajuputi seedling seed orchard in Indonesia were collected and air-dried, and their 1,8-cineole content and oil concentrations were determined. NIR spectra of the leaves were obtained, and calibrations for 1,8-cineole content and oil concentration were developed using spectra that had been selected using spectral features; that is, no knowledge of 1,8-cineole content or oil concentration was used to select the calibration samples. The calibrations were used to predict the 1,8-cineole content and oil concentration of the remaining samples. It was demonstrated that NIR spectroscopy could be used to identify leaf samples that had high 1,8-cineole contents and oil concentrations. The technique has the potential to greatly reduce the time involved in ranking large numbers of samples for these attributes, as is a requirement in tree breeding programs to enhance oil production.