The Australian Bureau of Statistics has conducted numerous surveys that are used to analyse poverty and the distribution of income amongst Australian households. Confidentialised unit record data for those surveys held since 1975 are available for the use and scrutiny of researchers. Recently, concerns have arisen over the reliability of these data to represent the circumstances of the population, and especially changes therein over time. This paper examines the quality and inter-temporal comparability of these survey data by comparing aggregates derived from the surveys to external data such as official population estimates, labour force data, the National Accounts and administrative data. We summarise the major changes to the survey data in an Appendix. Issues discussed include mis-reporting of income, and differences in scope, weighting procedures, definitions and collection methodology. The analysis suggests that uncritical use of the data may give rise to flawed estimates of the extent of poverty and inequality in Australia and how these have changed over time. There is scope to improve the comparability of the survey data, and the SPRC is pursuing this task in partnership with the ABS.