Correlation of disease phenotype with protein profile (proteotype) is a significant challenge for biomedical research. The main obstacles have been the need to insure sufficient quantities of pure protein sample, the reproducibility of protein display, and rapid and accurate protein identification. We present a modified approach that combines enhanced detection sensitivity with tissue microdissection from frozen primary renal cancer tissues of different histological subtypes, followed by 2D gel analysis and protein identification with MALDI mass spectrometry. We obtained reliable and highly consistent results in phenotypically similar tumors of each individual subtype by performing strict morphological control of the analyzed tumor cells without physical or chemical alteration of the frozen tissue samples. By application of non-oxidizing silver staining, proteins were resolved and identified with high levels of specificity and sensitivity. This new combination of techniques allows not only for sensitive identification of specific protein patterns that correspond to a histological tumor phenotype, but also for identification of specific disease-associated protein targets.