Although mutations in cell cycle regulators or spindle proteins can perturb chromosome segregation, the causes and consequences of spontaneous mitotic chromosome nondisjunction in human cells are not well understood. It has been assumed that nondisjunction of a chromosome during mitosis will yield two aneuploid daughter cells. Here we show that chromosome nondisjunction is tightly coupled to regulation of cytokinesis in human cell lines, such that nondisjunction results in the formation of tetraploid rather than aneuploid cells. We observed that spontaneously arising binucleated cells exhibited chromosome mis-segregation rates up to 166-fold higher than the overall mitotic population. Long-term imaging experiments indicated that most binucleated cells arose through a bipolar mitosis followed by regression of the cleavage furrow hours later. Nondisjunction occurred with high frequency in cells that became binucleated by furrow regression, but not in cells that completed cytokinesis to form two mononucleated cells. Our findings indicate that nondisjunction does not directly yield aneuploid cells, but rather tetraploid cells that may subsequently become aneuploid through further division. The coupling of spontaneous segregation errors to furrow regression provides a potential explanation for the prevalence of hyperdiploid chromosome number and centrosome amplification observed in many cancers.