To assess the functional significance of nuclear pore complexes, we have investigated whether the number of pores per nucleus is determined by such factors as the nuclear volume, nuclear surface area, DNA content, or aspects of nuclear activity. Comparisons were made between cell types chosen to permit observation of differences in nuclear pore number as a function of differences in the other qualities measured. The number of nuclear pores was determined by freeze-etching and measurements of nuclear surface and nuclear volume by electron and light microscopy. Pairs of cell strains in culture that contained different numbers of chromosome sets were investigated to examine the relation of pore number to total DNA content. Tetraploid cells of the rat kangaroo (Potorous tridactylus) have almost exactly twice the number of pores found in the parental diploid strain. However, the pore number in diploid grassfrog (Rana pipiens) cells was only 65% greater than in the parental haploid cells. In addition, a polyploid series of nucleated RBC had a 62% pore number increase with each successive increase in ploidy. Diploid cell strains from the canyon mouse (Peromyscus crinitus) and from the cactus mouse (P. eremicus) were compared to test whether a difference reflecting the 36% additional DNA in cells of the latter, associated with extra heterochromatin, existed. Although both were found to have the same number of pores and nuclear surface area, the cells differed in nuclear volume. These observations suggest that the number of nuclear pores is independent of the total amount of nuclear DNA, the nuclear surface area (and, thus, presumably the fraction of DNA that is bound to the nuclear membrane), the nuclear volume, and the size of the genome. Rather, the number of nuclear pores appears to be associated with some aspect of nuclear metabolic activity, e.g., transcriptional capacity or release of products to the cytoplasm. Further evidence for such a view comes from studies of chick embryo erythroblasts. In these, nuclear pore number was found to be lower in associated with the decreasing nuclear transcriptional activity and longer generation times that characterize the successive cell divisions leading to the fully differentiated state. The number of pore complexes reconstructed in the last cell cycles declined in a manner consistent with reutilization of previously formed pores in the absence of new pore synthesis. Challenging this interpretation is the increase in pore number at lower metabolic activity when Xenopus laevis cells are grown at different temperatures. The speculation that pore complexes have a longer half-life in cooler grown Xenopus cells could resolve the discrepancy.