Dividends have direct cash flow consequences for investors and are important for signalling reasons. Consequently, investors, analysts and managers typically forecast future dividends and report them in various ways. Yet the accuracy of dividend forecasts has been largely neglected in empirical finance. We examine the accuracy of managers' dividend forecasts in Australian IPO prospectuses (a companion paper examines the analysts' dividend forecasts). Managers' dividend forecasts are optimistically biased. Nevertheless, they are substantially more accurate and less biased than their earnings counterparts. Differences in retained ownership and the predictability of earnings help explain why some dividend forecasts are more accurate than other...