<b> Abstract <sub><br> <i> This article is fundamentally a dataset required for analysing some taxation variables. The data describes the raw, combined, and filtered tax information on company-level such as Tax planning, Tax Saving and Tax disclosure in Malaysia. It combines a firm-level panel dataset and seven variables: tax disclosure, tax planning and a component of tax saving (namely: permanent differences, temporary differences, foreign tax rates, tax losses or NOLS and unclassified ETR reconciling). The dataset results from 2,811 company-year observations for the period 2010 to 2012, excluding financial companies (195 Obs), firms whose data needed is not available (48 Obs), firms whose date of fiscal year-end is not excluding 31/12 and Accounting period is more than 12 months (495 Obs), loss-making companies at least in one of the years (894 Obs), firms their annual reports not available for three years (291 Obs), firms with an extreme value of ETR (18 Obs) and firms with outliers and influential observations (12 Obs), resulting in a final sample of 858 company-year observations. The data were extracted from company annual reports. Researchers can smoothly use this dataset as two data types were created to allow individual and compound measurements. The individual determinant is the direct measurement of tax planning. For the compound measurements for tax disclosure, we used the index of ETR reconciling items and the five components of tax saving as undirect measurements of tax planning.