This paper discusses the literature that shows that declaring a zero vision for everything bad (including unsafe behaviours, incidents, injuries) does not prevent fatalities or major accidents. In fact, parts of the literature show that a reduction in minor badness increases the risk of major accidents and fatalities. This is true in several industries. Two families of explanations are discussed. The first is the concern that declaring a zero vision can reduce operational knowledge. The second is the unsubstantiated assumption that minor injuries and fatalities have the same causal pattern. In general, evidence for or against the utility of a zero vision is dogged by confounding factors (other variables responsible for changes in safety outcomes) and what Giddens called the double hermeneutic, where the results of such studies are only as stable as the attributions the original reporter (e.g. OHS official, case worker) and the subsequent analyst (e.g. researcher) made about a particular event. The paper concludes that in a complex, dynamic, resource-constrained and goal-conflicted world, zero is not an achievable target, but a zero commitment may be worth some encouragement.