The economics of controlling house mice using 0.005% bromadiolone-treated wheat placed in bait stations within soybean crops was evaluated when mouse densities were approximately 200-300 per hectare. This method of control was available to soybean growers in New South Wales during 1989. Damage by mice to soybean crops (n = 4) was assessed up to 200 m into two crops (one with bait and one without) on each of two farms just before baiting, and 22 and 42 days after baiting. Mouse damage in each crop was always less than 10%, and was greatest in patches of high crop density. Damage was similar at the crop perimeters and up to 100 m into the crops, but 200 m from crop perimeters, damage was less prevalent. By 42 days after baiting, the differences in crop losses relative to pre-baiting damage levels were minimal. Thus the economic benefits of using bromadiolone in bait stations, were, at best, marginal. Estimates of losses caused by mice in three of the principal soybean growing regions of New South Wales were obtained via a questionnaire distributed by mail to all farmers who grew soybeans in 1988-89. These data indicated that the combined losses caused by mice to growers in these resions in 1989 was about $A2 million.