Productivity indices, developed by the Australian Greenhouse Office as part of the National Carbon Accounting System (NCAS) to aid spatial estimation of native forest productivity across Australia, were tested to estimate site quality for Pinus radiata D.Don plantations at farm level in and around the tablelands in the south-eastern corner of New South Wales. This study found a significant (P < 0.01) relationship between site index and pro-ductivity indices generated by NCAS. 'Region', included as a categorical variable, was significant (P < 0.05) in the relationship between site index and these productivity indices. These results were consistent between indices that used long-term average data and those that used only data that coincided with the lifetime of the plantations. Various groupings of regions were tested to produce a logical and useful model to estimate site index. However, the causes of the significant differences between regions are not known and the classification of 'Region' is speculative for unplanted areas. The paper discusses differences in the models that use long-term or short-term productivity indices and the potential to transfer productivity estimates derived at national scale to local scale.