Based on two household surveys, this article investigates the over-time evolution of the determinants of living standards in Azerbaijan during the transition from a centrally planned economy to a market economy. Quintile regression is used to estimate the relative importance of inter-temporal changes in individual, household, and environmental factors to household expenditures from 1995 to 2002. Results highlight systematic differences in the effect of some factors over the investigated period. Some tendencies uncovered at the beginning of the reforms – for instance, the strong positive effect of renting land – disappeared completely as reforms progressed over time. In contrast, several new factors – such as the strong positive effect of a young age of the household head, and the strong negative effect of having more women in the household – emerged over time as new significant predictors of living standards. Yet, the influence of other factors – for example, the strong negative effect of having a large number of dependents in the household – remained fairly stable inter-temporally. In addition, three variables – namely, university education, a large number of dependents in the household, and location – were found to be specifically important for improving the living standards of households in the poorest quintile. These findings suggest that the emphasis of poverty reduction policies in Azerbaijan should focus on improving the quality of, and access to, university education, developing an effective social protection system, and reducing regional disparities.