New techniques for the early detection of cancer are fast emerging. This is essential for more effective diagnosis and control of the disease. We have used a High Performance Liquid Chromatography-Laser Induced Fluorescence (HPLCLIF) technique to record chromatograms of proteins in serum and ovarian tissue samples. The recorded chromatograms of normal, benign and malignant samples were analyzed using statistical (Principal Component Analysis) methods. It is shown that chromatograms of the samples can be classified into sets, and a model based on such a classification can be used to analyze protein profiles of test samples of serum and ovarian tissue for the detection of malignancies.