ABSTRACT We examine the properties of firms' forecasting records and whether the accuracy of their prior earnings forecasts affects investor response to their subsequent forecasts.Within the context of a Bayesian model of investor learning, we find that the stock price response to management forecast news is increasing in prior forecast accuracy and also in the length of a firm's forecasting record. Further, we document that investors are more responsive to extreme good and bad news forecasts when a firm has an established forecasting record. Overall, these results suggest that a firm's prior forecasting behavior allows it to establish a forecasting reputation, and that market forces encourage accurate forecasting as firms benefit from having a reputation for forecasting accurately. Data Availability: Data are available from public sources cited in the text. JEL Classifications: G19; G39; D89; M40.