Theories suggest that corporate governance mechanisms affect corporate dividend policies. This study extends and tests the implications of two extant static agency models making opposite predictions. The outcome model predicts an increase in dividends when the corporate governance mechanisms improve, because shareholders are better able to force managers to disgorge cash. In contrast, the substitute model suggests that an improvement in the corporate governance mechanisms reduces the role of dividends in controlling agency costs, leading to a decrease in dividends. This paper investigates the dividend policy for firms listed on Saudi Arabia Stock Exchange. This is a case study of Saudi Stock Market, where the determinants of dividend policy have received little attention. This study use a panel dataset of non-financial firms listed on Saudi Arabia Stock Exchange between the years of 2007 and 2010. Based on a panel of 366 firm year observations of 99 Saudi firms, we provide evidence in outcome model or substitute model with ownership structure, board structure and debt policy. Three Tobit models are specified: In the first, we construct a governance index based on eight criteria: seven criteria which capture various aspects of a firm’s structure, policies and practices that constitute good governance and a criterion that examines the company’s compliance with Shariah law in all its activities. Therefore, we estimate the effect of corporate governance on dividend policy in the first model. In the second, we investigate how dividends interact with corporate governance mechanisms in a panel of data. We explore the relation between dividends and ownership structure (ownership concentration and managerial ownership), board structure (board size, Board independence and Chairman-CEO duality) and debt policy. In the final, another test of the substitute and the outcome models is built on the Jensen (1986) free cash flow theory, which states that dividend policy can extract surplus cash from management control by reducing free cash flow. In this third model, we examine how corporate governance improvements affect the dividends’ sensitivity to free cash flows by focusing on the coefficients on the interactive variables of the ownership structure, board structure, debt policy and the free cash flow. For the three models, we divide sample in two subsamples and we compare the results obtained by using criteria of company’s compliance with Shariah law. For the effects of corporate governance (measured by corporate governance score) on dividend levels, we find that dividend policy is a substitute model for good governance for all Saudi Arabia firms. When we select only Shariah compliant firms, results indicate also that dividend policy is a substitute model for good governance but results are insignificant. When we select only Non-Shariah compliant firms, results indicate the same conclusion. We find that governance is associated with fewer dividends, supporting the substitute model and indicating the influence of good governance by forcing less cash to be returned to investors. For the effects of corporate governance mechanisms on dividend levels, we find that the only variable affect the dividend levels for Non-Shariah compliant firms is the separation in the functions of chairman and of CEO supporting the substitute model. For Shariah compliant firms, dividend policy is an outcome for the separation in the functions of chairman and of CEO, and ownership concentration. Governance through the separation in the functions of chairman and of CEO and ownership concentration influences firms by forcing more cash to be returned to investors. For the effects of the corporate governance improvements on dividends’ sensitivity to free cash flow, our results support the substitute hypothesis for Shariah compliant firms regardless the board independence, board meeting, managerial ownership and debt. Improvements in these corporate governance mechanisms reduce firms’ need to force out the free cash flow through dividends. For Non-Shariah compliant firms, our results support the outcome model for managerial ownership and ownership concentration