We review the principles and applications of statistical filtering in multichannel fluorescence microscopy. This alternative approach to separation of signals from individual fluorophore populations has many important advantages, especially when spectral and/or temporal overlap, or the complicated nature of those signals, makes their discrimination or sorting impossible by means of hardware. This situation is typically encountered for biological samples. This review of well established statistical filtering techniques and of emerging, very promising new methods of analysis reveals remarkable progress in bioanalytical applications of fluorescence microscopy.