This study provides a proof of concept of using remote sensing to classify species of small-scale plantation at a regional level and achieved high classification accuracies for most species. Douglas-fir and eucalyptus were the two most accurately classified alternative species, with over 90% of producer’s accuracy. The most important input variable selected for the classification was DEM (Digital Elevation Model), suggesting that elevation plays an important role in differentiating plantation species. The accuracy of species classification highly depends on the availability of truthing data. In total, 2151 ha of alternative species were classified for Hawke’s Bay and a majority of them are eucalyptus, cypress and poplar. The transferability of classification derived from one region to another region is low due to regional variations in the topography, climate and species composition. In order to map the national cover of alternative species, truthing data that cover a range of species and ages classes from all regions are required. One limitation with the study is that pre-defining the geographic boundaries of alternative species is required to define the extent of classification, as the current small-scale plantation map developed by the School of Forestry may not pick up all the alternative species. Without the pre-defined boundaries, the classification approach tends to map other land covers as alternative species plantations due to a similar spectral signature.