In this review, I stress the importance of direct data and accurate terminology when formulating and communicating conclusions on how the G2/M and metaphase/anaphase transitions are regulated. I argue that entry into mitosis (i.e., the G2/M transition) is guarded by several checkpoint control pathways that lose their ability to delay or stop further cell cycle progression once the cell becomes committed to divide, which in vertebrates occurs in the late stages of chromosome condensation. After this commitment, progress through mitosis is then mediated by a single Mad/Bub-based checkpoint that delays chromatid separation, and exit from mitosis (i.e., completion of the cell cycle) in the presence of unattached kinetochores. When cells cannot satisfy the mitotic checkpoint, e.g., when in concentrations of spindle poisons that prohibit the stable attachment of all kinetochores, they are delayed in mitosis for many hours. In normal cells, the duration of this delay depends on the organism and ranges from ∼4 h in rodents to ∼22 h in humans. Recent live cell studies reveal that under this condition, many cancer cells (including HeLa and U2OS) die in mitosis by apoptosis within ∼24 h, which implies that biochemical studies on cancer cell populations harvested in mitosis after a prolonged mitotic arrest are contaminated with dead or dying cells.