Yes, but only for those firms that are highly likely to pay dividends. Using U.S. data over the period 1970-2005, I investigate the value effect of dividends. To control for the potential selfselection bias, I examine the relation between firm value and dividends among firms with similar likelihoods of paying dividends. I find that the value of dividend-paying firms is not consistently higher or lower than the value of non-dividend-paying firms. In general, firm value does not increase or decrease with the amount of dividends for the majority of firms. However, there is evidence that high dividends increase firm value among firms for which dividend payouts are highly likely. Overall, I find little empirical support for the simple tax and free cash flow hypotheses. My evidence indicates the presence of a dividend clientele that favors high dividends. More importantly, it appears that this dividend clientele factor in not just high dividends but also key firm characteristics?such as stability of cash flows, maturity, profitability, and firm size?into investment decisions.