Several studies have shown that patterns of gene expression remain consistent with the tissue of origin in cancer samples. Gene expression profiling may therefore offer a promising new technology to build a site of origin classifier with the ultimate aim to determine the origin of cancer of unknown primary(CUP). A single cDNA microarray platform was used to profile 229 tumors of known origin(14 tumor types). This data set was subsequently used for training and validation of a support vector machine(SVM)classifier, demonstrating 89% accuracy to predict a site of origin(13 types). Applying this microarray SVM classifier to 13 cases of CUP, a high confidence prediction was made in 11 of 13 cases. These predictions were supported by comprehensive review of the patients' clinical histories. Thus, data generated using both microarray and quantitative PCR can be used to train and validate a cross-platform SVM model with high prediction accuracy.