Purpose: The objective of this paper is to present the dividend signaling hypothesis, in particular, an empirical analysis of the relationship between the current changes in the level of a dividend paid (t0) and the future profitability of the companies. Design/Methodology/Approach: The dividend signaling hypothesis is empirically tested using the dynamic causality analysis based on the regression approach monitoring for expected earnings changes and past returns with a set of linear and non-linear controls. The conducted analyses comprised the domestic companies quoted on the Warsaw Stock Exchange, which paid dividends in 2001-2016. Findings: The empirical results confirm that in the audited period, dividend decisions bring some information about the current situation (t = 0) and future (t = 1, t = 2) of the analyzed companies. It is also worth noting that among the analyzed indicators, the gross profit ratio (PBT) referred either to the market value or the book value of equity was most often in the statistically significant analyzes. In general, our results confirm the validity of the signaling hypothesis in the case of continuation-growth and initiation of payments in the Polish capital market as in the developed market. Practical Implications: We can say that investors, based on "signals" coming out of dividends advertisements, may conclude the future-income potential of a given company. Originality/value: As mentioned, the gross profit ratio (PBT) referred either to the market value or the book value of equity was most often in the statistically significant analyses. Thus, it seems that, contrary to the adopted assumptions in the literature, gross profit (PBT) in dividend signaling is more important than net profit (E).