The potentialities of polarization microscopy has been greatly increased by using specific stains for selective enhancement of the optical anisotropy of a macromolecular constituent of cells and tissues. Such stainings have proved to be especially useful in exploring the spatial orientation pattern of the extracellular matrix components. The retardation value, which characterizes quantitatively the degree of submicroscopic orientation, can be measured traditionally with a compensator plate. This technique, however, is time-consuming and greatly dependent on visual judgment. Several attempts have been made to combine digital image analysis and polarization microscopy to improve the measuring technique in unstained structures. In this paper, we summarize theoretical considerations and experimental data to show the advantages and limitations of this methodological approach when using stained and birefringent specimens. The technique we are suggesting is the measurement of the light intensity using a 12 bit cCCD camera attached to a polarized light microscope and digital image analysis system. The theoretical basis is given by the Fresnel equation describing the relationship between light intensity and retardation value. According to this, there is a sin2 function between the light intensity and the retardation value. The same relationship of these two parameters was observed in our experiments on the birefringent extracellular matrix around chondrocytes grown in agarose gel and interterritorial and territorial matrix of canine articular cartilage stained with picrosirius red. Our results suggest that the retardation values can be calculated directly from the light intensity values if the retardation value is lower than lambda/2.