The hypothesis that corporate directors determine dividend payments by applying target payout ratios and related parameters to current and past earnings has gained widespread acceptance in recent years as a generally applicable account of dividend policy. One reason for the acceptance of this view appears to be its success in predicting the behavior of aggregate measures of dividends in time series regression analyses. The interpretation of these findings as supporting the target payout hypothesis may well be incorrect since the behavior of aggregate magnitudes need bear no direct and close relationship to the behavior of individual firms. This paper attacks the target payout hypothesis on two grounds: it shows that the behavioral characteristics of the firm assumed by the hypothesis have no satisfactory basis in managerial motivation; and it presents data relating to the dividend policies of individual firms for which the hypothesis fails to account.