We have examined a variety of conditions for solubilizing and electrophoresing cell proteins in order to define optimum conditions for studying proteins modified by ADP-ribosylation. We have identified conditions in which proteins can be quantitatively extracted from cells in an undegraded form with the protein-ADPribose linkages intact. Effective measures include boiling cells briefly (4 min) in the presence of 2% SDS and 2 M urea at pH 6.8. Both SDS and urea were present in the 6–18% gradient polyacrylamide gel matrix used for electrophoresis. Under these conditions good resolution of proteins of a wide molecular-weight range is obtained. This system has been used to compare protein ADP-ribosylation in non-transformed and polyma virus-transformed baby hamster kidney (BHK) fibroblasts, since the latter cells have a greater NAD + ADP-ribosyltransferase activity (measured in isolated nuclei and permeabilized cells). Addition of DNAase to permeabilized BHK cells over the range 10–150 μg led to a progressively greater activation of transferase compared with controls. When PyY cells were used, however, maximum activation was achieved with only 10 μg of DNAase, further additions producing a successively smaller activation relative to control cells without added nuclease. There were also differences between these cells in response to salt. Addition of NaCl (to about 0.3 M) to BHK cells resulted in various extents of transferase activation, whereas any addition of NaCl to the incubate of permeabilized PyY cells decreased transferase activity. These different enzyme activities between this transformed and non-transformed cell line are for the most part not reflected in the protein modification profiles seen on autoradiograms of acrylamide gels after electrophoresis of 32P-labelled proteins. A variety of proteins are modified and their molecular weights depend on the NAD concentration in the permeabilized cell incubation. At 0.5 μM NAD + there were two major acceptors with M r values of 14 kDa and 30 kDa, and at 100 μM NAD +, three major acceptors, with M r values of 19 kDa, 45 kDa and greater than 170 kDa. NAD concentrations of between 1 μM and 100 μM had no further effect on protein ADP-ribosylation profiles, except for the protein(s) of M r > 170 kDa, pointing to a critical difference around 0.5–1.0 μM substrate. In some experiments, however, a difference was observed in the intensity of radioactivity in two bands. This may represent two different proteins, or a single protein modified to different extents. Since the molecular mass values were less than 12 kDA, we presume the proteins (or protein) are non-histones. Kinetics of NAD + utilization show an initial burst of activity. Autoradiographs show little further change in protein labelling pattern or intensities after the first 10 s of an incubation. The overall rate of ADPribose degradation is slow in these permeabilized cells, only about 30–35% being lost in a 90 min degradation assay. Differences are seen in the autoradiographs, however, between the two cell types when 32P-labelled proteins, solubilized at different times during the degradation assay, are electrophoresed in the polyacrylamide gels. Finally, 3-aminobenzamide takes about 15 min to completely inhibit ADPribose synthesis in the permeabilized cells. During this time, synthesis is seen from autoradiographs to be limited to the modification of three particular proteins of molecular mass 25 kDa, 54 kDa and 170 kDa.