Drugs targeting the mitotic spindle are used extensively during chemotherapy, but surprisingly, little is known about how they kill tumor cells. This is largely because many of the population-based approaches are indirect and lead to vague and confusing interpretations. Here, we use a high-throughput automated time-lapse light microscopy approach to systematically analyze over 10,000 single cells from 15 cell lines in response to three different classes of antimitotic drug. We show that the variation in cell behavior is far greater than previously recognized, with cells within any given line exhibiting multiple fates. We present data supporting a model wherein cell fate is dictated by two competing networks, one involving caspase activation, the other protecting cyclin B1 from degradation.