Australian research on poverty and income distribution has been subject to criticism over definitional, data quality and measurement issues that have undermined its validity and impact. In reviewing these issues, this paper addresses some of the problems that have been identified in the literature and illustrates their importance with examples that shed light on the validity of the arguments. It also seeks to establish whether poverty has increased or decreased in the decade to 2002–03 and examines the overlap between hardship or deprivation and income poverty, and reviews some of the emerging evidence on the dynamics of low income. After a detailed review of data problems, the significance of definitional issues is illustrated in relation to the measurement of income inequality and poverty. Attention then focuses on assessing the sensitivity of estimated trends in income poverty to how poverty is defined and measured. Although there is variation across the different measures, the evidence suggests that progress in reducing relative poverty has been slow. Our preferred measure (based on current income) suggests that relative poverty increased slightly over the decade, with most of the increase occurring between 1995–96 and 1999–2000. When poverty is measured in ‘absolute’ terms using a poverty line that is held constant in real terms, the poverty rate declined overall, but was more or less constant between 1996–97 and 2000–01. Estimates derived from wave III of Household Income and Labour Dynamics in Australia (HILDA) data indicate that those who were income poor in 2002–03 experienced twice as many hardship conditions as those who were not poor, with around one-quarter of the poor having problems paying bills or seeking external financial assistance. HILDA data also indicate that many of those who are in poverty in any single year manage to escape within the next 2 years, although it is too early to know whether they escape permanently, or slip back into poverty or oscillate on its margins.