Recent oil labeling around reduced trans fat levels, have increased the interest in naturally stable oils.  DAS has developed the Natreon^TM canola oil (Brassica napus variety Nexera) to address this market. A non-destructive, single seed biochemical screen was developed to further improve the currently available Natreon^TM canola oils.  This screen is based on FTNIR prediction models to measure the main seed components such as oil and protein.  Application of this high throughput analytical method to screening large canola populations allows us to accelerate selection of canola with improved trait attributes. A commercially available liquid handler was modified to generate vacuum for the purpose of manipulating individual seeds between custom made envelope racks and an integrating sphere.  Custom software was developed to randomly select single seeds from a seed envelope or a 96 well microtiter plate by detecting vacuum changes. The seed is scanned using a commercially available FTNIR equipped with a fiber optic probe mounted on a small Spectralon integrated sphere.  The seed is maintained inside of the sphere by the vacuum at the tip of a fixed canula while the spectrum is collected.  Several FTNIR multivariate prediction models have been developed.  The models are based on single seed FTNIR measurement and the measured reference values representing the particular seed components.  These reference values were generated by DAS developed chromatographic techniques using single seed extracts. As of today, prediction models have been developed for oil content, various fatty acids, protein content, seed weight, and minor (<5%) biochemical components.This talk will describe the automated HT system as well as present biochemical results.