Purpose: This study looks at the relevance between discretionary revenue and book-tax differences (hereafter BTDs). While the study of earnings management, which focused on discretionary accruals and real earnings management, has largely made, it has not yet been actively researched on discretionary revenues. Therefore, it was believed that discretionary revenue would expand the preceding study by looking at its relevance to BTD, known as financial reporting quality and measures of tax avoidance. In general, prior research suggested that earnings management make BTDs larger. Thus, the relationship between discretionary revenue and the amount of BTD is predicted positive. Research design, data and methodology: To this end, the method of discretionary revenues was used and BTDs measured in four ways. First, Earnings before income tax – estimated taxable income divided by total asset (BTD). Second is fractional rank variable of BTDs (FBTD). Third is Indicator variable equals 1 if the firm-year has a positive BTD, 0 otherwise (PBTD). Fourth is that Indicator variable equals 1 if the firm-year has a BTDs in top(bottom) quartile, 0 otherwise (LPBTD, LNBTD). 4,251 samples were analyzed in the Korean Security market (KOSPI) from 2003 to 2014. Results Empirical analysis shows that BTDs increases as discretionary revenue increases. These results were equally observed when BTDs was measured as a ranking variable or as a indicating variable. These results indicate that earnings management through the revenue of managers exacerbate the quality of financial reporting. Conclusions: In sum, discretionary revenues can be used as an indicator of making BTDs larger and meaningful as the first study of the Korean capital market where discretionary revenues affect accounting information quality. Investors need to increase interest in discretionary revenues because intervention in financial reporting through revenue accounts by managers can increase information asymmetry and agency costs. This means that studies on discretionary revenues that have been relatively small should be expanded. The results also provide important implications for the relevant authorities and investors. Despite these benefits, however, measurement error problems with estimates still appear as limited points, and prudent interpretations are required, and additional follow-up studies are needed in that variables that are not yet considered in this study may affect our findings.