Often, measurement of biological components generates results, that are corrupted by noise. Noise can be caused by various factors like the detectors themselves, sample properties or also the process of data processing and appears independently from the applied technology. When measuring two identical samples it can be observed that similar signal intensities may have inherent but varying levels of noise and that the ratio of noise decreases with increasing signal intensities. In this paper a statistical approach is introduced to estimate the noise inherent in the measured data. Based on this estimation, it is possible to provide information about the reliability of a measured signal and whether the difference between intensities is mainly caused by noise or by biological relevant cellular alterations.