Cancer has become a serious health problem worldwide in the last decades based on the increasing cancer incidences and death cases. National statistics (2010) show, that cancer is ranked as the second most common cause of death in Austria and led to almost one out of four death cases in men. Prostate cancer in particular is the most frequent diagnosed cancer type and caused almost every tenth cancer-related death since 1994. The early detection of a disease can have a great influence on the progress and outcome of the disease. The most reported tumors of the prostate gland occur later in life of men, and detection is often difficult as they are asymptomatic and painless. Due to the mainly slow development of prostate cancer, new diagnostic approaches for early diagnosis could improve the detection rate and curability of prostate cancer. One function of the prostate gland also is the secretion of the prostate specific antigen (PSA) in the urethra and the circulation. Diminutive amounts of PSA can be detected in the blood, the PSA-value, and can be used as a prostate specific biomarker. Due to increment of the PSA production during cancer development, the screening of the PSA-value is widely used as a diagnostic tool to detect or confirm prostate cancer. Based on the fact, that a variety of reasons can cause an increased PSA-value, the PSA-screening led to a high false positive rate. A promising diagnostic approach could be the use of the protein microarray technology to identify specific biomarker which can differentiate between prostate cancer and control samples. Due to the high-throughput ability of the protein microarray technology, thousands of antibodies or tumor associated antigens (TAAs) can be screened in one single experiment. These TAAs are elicited in the circulation during a tumor development and led to the generation of tumor-autoantibodies as a result of an immune response. Different research groups are focusing on the identification of a panel of prostate specific autoantibodies to apply them as an early detection method using blood samples. Therefore, proteins of the well defined in-frame clones of the UNIPEX- library were expressed and spotted as duplicates onto microarray slides. The immunoglobulin G (IgG) of 100 plasma samples, subdivided in 50 prostate cancer and 50 control samples, was isolated and incubated on the non-processed microarray slides containing almost 15,300 proteins. Image- and statistical analyses were applied to the collated data. The performed 1-Neares Neighbor classifier, using a significance level of 0.001 and the quantile-normalization resulted in the identification of 167 classifiers in the IgG purified samples with an accuracy of 92 %. The cross-validation from the Bayesian compound covariate predictor distinguished between the prostate cancer and control samples with an ROC curve AUC of 0.838. Due to the high significance of the AUC, the identified classifiers could be used as candidates for the targeted array.