Development of predictive formulae or qualitative statements about the probable outcome of control campaigns entails knowledge of bait quality and distribution, behaviour, vagaries of weather at the time of the control campaign, and other factors which govern the probability that the target animals will accept bait. This paper collates experience in recognising, estimating and using some of these variables for predicting the outcome of large-scale poisoning, and discusses possible approaches to the solution of some hard-core problems.

Major causes of failure of aerial poisoning operations against possums identified were: sublethal toxic loading, undersize sublethal baits, nonlearned behavioural aversion to 1080, and failure to encounter bait. Dislike of bait was not a major cause of failure. Progress has been made towards solving these problems, but failure to encounter bait remains a likely major reason for possums surviving aerial poisoning. Improvements in the aerial sowing of bait are essential if the full benefit of this progress is to be realised.