A simple illustration of different potential causes for leakage of toxic waste containers in a depository illustrates how knowledge of a leakage event can be used to update the probability that one basic cause is more likely than another. The method invoked for this analysis is that of Bayesian probability. The point of doing the work is to show how one can come to grips with the most likely reasons for such a leakage. In turn, such improved knowledge, gleaned from the actual leakage events that occur, allows one to better design options for remediating the depository site based on using the potential causes of leakage and their improved likelihood of being correct. The example shows how one can decide between different leakage causes based on measurement of particular leakage events.