A new procedure for analysis of random telegraph signals in time domain has been developed and applied to the analysis of voltage fluctuations in the current induced dissipative state in superconducting thin films. The procedure, based entirely on the difference in the statistical properties of discrete Marcovian telegraph fluctuations and Gaussian background noise, ascribes each point of the experimental time record to one of the telegraph states. The average statistical lifetimes and amplitudes of the telegraph signal are then determined in an iterative way by fitting the amplitude histogram of thus obtained record of the redistributed data to the two-Gaussian histogram of the original experimental signal. The procedure allows for analyzing “noisy” random telegraph signals with low ratio between the signal amplitude and the intensity of the background noise that cannot be analyzed by the classical approach. Separation of the time record into two subrecords relative to two telegraph states also enables in-depth analysis of the spectral properties of the background noise observed together with the telegraph fluctuations.