Modelling moth dispersal in relation to wind direction and strength could greatly enhance the role of pheromone traps in biosecurity and pest management applications. Anemotaxis theory, which describes moth behaviour in the presence of a pheromone plume and is used as a framework for such models. Currently, however, that theory includes only three components: upwind, zigzagging, and sideways casting behaviour. We test anemotaxis theory by analysing the data from a series of mark–release–recapture experiments where the wind direction was known and the insects were trapped using an irregular grid of pheromone traps. The trapping results provide evidence of a downwind component to the flight patterns of the released insects. This active or passive downwind dispersal is likely to be an appetitive behaviour, occurring prior to the elicitation of pheromone-oriented flight patterns (pheromone anemotaxis). Given the potential for significant displacement during downwind dispersal, this component will have impact on final trap captures and should be considered when constructing moth dispersal models.