Predator-prey theory suggests that pest damage that occurs over a period of time is related to the initial pest density and their functional and numerical responses to the resource. Estimating the relationship between damage and initial density is therefore fundamental to determining the threshold density (DT), above which the economic benefits of control exceed the economic costs of control. A manipulative experiment was conducted to parameterize a model of DT for managing yield loss (YL) caused by mice (Mus domesticus L.) from the milky stage to harvest, by assessing the relationship between YL and initial mouse density at the milky stage (DI). This relationship subsumed the functional and numerical response of mice to maturing wheat crops. The experiment was conducted in a typical wheat field and DI was manipulated by releasing a known number of mice into an experimental crop enclosed by a mouse proofed fence. The estimated relationship was asymptotic exponential: YL increased almost linearly with DI until apparent competition between mice occurred at densities in excess of about 500 mice ha1; and apparent competition limited further increases in YL at densities in excess of 2133 mice ha1. DT varied depending on the effectiveness of the control method in reducing DI and the cost of control as a percentage of the farm-gate value of wheat. DT was 89 mice ha1 for triggering aerial baiting if this method was effective in reducing DI by 41% and if the cost of baiting was 7.54% of farm-gate value of wheat. Grain growers will be able to use DI immediately to make informed, timely decisions for managing mouse damage to wheat crops. Mouse densities should be monitored well before the milky stage so that DI can be forecast by the trend of monitored densities. If the forecast DI exceeds DT, management should be implemented at least a week before the milky stage (allowing a week for control such as baiting to take effect). However, if the forecast DI is much higher than D
