Technical Abstract: Genetic and environmental effects on carbon allocation to soluble phenolics and non-structural carbohydrates in needles of widely-planted loblolly pine (Pinus taeda L.) genotypes could impact productivity, sustainability and biogeochemical cycling in the southeastern U.S. The magnitude of genetic and environmental influences on needle chemistry, however, is not well-characterized. The objective of this study was to investigate genetic and site-quality effects on foliar concentrations of total phenolics, condensed tannins (i.e. proanthocyanidins, PA), starch and soluble sugars in two different field studies. In the first study, we measured total phenolics, PA and non-structural carbohydrate (NSC) in nine different genotypes representing a gradient of genetic diversity: three half-sib families, three full-sib families and three clones, growing in a plantation setting for two years. In the second study, we measured concentrations of soluble phenolics and PA in eight clones with different growth potentials after nine years of growth in a replicated clonal trial planted at two sites with differing soil quality. In the first study, differences in productivity, total phenolics, PA and NSC among the various genotypes were generally not statistically significant. Weak negative relationships between individual-tree volume, total phenolics and PA concentrations were observed, while soluble sugar concentrations tended to increase with tree volume. Starch concentrations among genotypes were highly variable. In the second study, there was a significant site × clone effect for total phenolics. Clones showing lower volume production had 21 - 23% higher mean total phenolic concentrations at the less productive site relative to the more productive site. More productive clones showed little variation in mean total phenolic concentrations across sites, suggesting that more productive clones may exhibit less plasticity in allocation to these compounds. However, we found a positive relationship between clone mean total phenolics and mean tree volume at the more productive site, but not at the less productive site. Overall, these results indicate that the relationship between foliar concentrations of phenolics and productivity can be affected by both genotype and site. The relationship between productivity and phenolic concentrations was negative between sites, suggesting that mineral resource limitations at the low soil quality site promoted carbon allocation to phenolics, in agreement with carbon-nutrient balance theory. By comparison, phenolic concentrations at the high soil quality site were positively related to tree volume, suggesting that with increased resource availability, genetic potential for improved growth includes increased allocation to these secondary metabolites. Investigation of the underlying physiological and genetic factors that influence production of secondary metabolites may improve our ability to predict how deployment of highly-productive and widely-planted loblolly pine genotypes will affect pine plantation ecosystem dynamics.