Multicategory Support Vector Machines (MC-SVM) are powerful classification systems with excellent performance in a variety of biological classification problems. However, the process of generating models in traditional multicategory support vector machines is very time-consuming, especially for large datasets. In this paper, parallel multicategory support vector machines (PMC-SVM) have been developed based on the sequential minimum optimization-type decomposition methods for support vector machines (SMO-SVM). It was implemented in parallel using MPI and C++ on both shared memory supercomputer and Linux clusters, and used for multicategory classification. The performance of PMCSVM has been analyzed and evaluated using several datasets including two microarray datasets with totally 31 diagnostic categories, 25 cancer types and 12 normal tissue types. The experiments show that the PMC-SVM can significantly improve the performance of classification without loss of accuracy, compared with previous work.