Measuring effective tax rates using tax revenue data is attractive, given that revenues collected capture the net effect of tax provisions and taxpayer behaviour that are difficult to model. Yet reliance on aggregate tax and income data requires restrictive assumptions and significantly limits the scope of analysis. This paper considers advantages of relying on micro-data gathered from tax returns to assess average tax rates on labour, capital and transfer income and presents some illustrative results. The analysis emphasises the importance of matching taxpayerlevel information to income flows, and notes difficulties in interpreting tax rates that average over all taxpayers. It also illustrates the importance of loss adjustments in measuring effective tax rates on capital income, and reports evidence of significant variation in corporate average tax rates by sector and firm size.