Several nuclear and surface proteins are expressed in varying amounts in the different phases of the cell cycle. For some of them the coding gene is not known and changes in their expression could simply be secondary to changes in the proliferative activity of the population. Other proteins are oncogene products, probably having a direct regulatory function in cell proliferation, differentiation and malignant transformation. Studying these proteins may both permit a better understanding of the mechanisms regulating proliferation and differentiation and provide kinetic parameters for describing the cell cycle. Based on antibodies against these proteins, bivariate flow cytometry (FCM) is able to quantitate their expression simultaneously with DNA distribution. This allows protein expression to be related precisely with each cell cycle phase in populations having different proliferative activity. Further advantages of bivariate FCM are that few cells are required for the analysis and the percentage of cells expressing the (onco) gene product can be determined. Several cellular proteins have been investigated with bivariate FCM, and the data are reviewed. Some proteins not coded by oncogenes (such as cyclin, the Ki-67 reactive antigen and DNA polymerase alpha) are expressed in cycling, but not in G0 cells and are of special interest for the kineticist, since they could identify cells which are able to initiate DNA synthesis, i.e. those representing the "growth fraction" of the population. Statin, on the contrary, is apparently expressed only in G0 cells. The expression of some proteins coded by oncogenes, such as p53 and the c-myc product is high in proliferating G1 cells and decreases with differentiation. The expression of the c-ras product is not strictly related to cell cycle phases and increases with differentiation. Technical improvements (allowing, for example, the monitoring of the changes in protein expression following the microinjection of a protein-blocking substance into the cells and the inclusion of phenotype markers into the analysis) will expand the role of bivariate FCM for these research works.