The underlying molecular mechanisms of renal cell carcinoma (RCC) are poorly understood and more reliable markers for early diagnosis are needed. Hence, alternative strategies for biomarker discovery with appropriate validation technologies have to be performed. To elucidate genesis and progression of RCC we used high parallel chip based gene expression profiling comparing normal and tumour tissues. We compared corresponding control and tumour tissue samples from 10 patients with clear cell RCC. We isolated RNA from histologically well characterised tissue sections and performed reverse transcription, labelling and linear RNA amplification. Samples were hybridised on microarrays containing 642 human cDNAs. Of the 352 differentially expressed genes found, CD70 and FRA2 were selected for further evaluation by real-time RT-PCR. The analysis all showed a high potential to discriminate between normal and tumour tissue. Moreover, increased CD70 mRNA expression in tumour cells could be correlated to its expression at the protein level. Immunohistochemistry (IHC) showed very strong expression of CD70 in all tumour samples but no expression in adjacent normal kidney tissue. With our combined approach we were able to identify CD70 as a new marker for RCC, which may be useful in the future for improved immunohistochemical diagnosis.