Three issues are addressed in this paper. First, we use both household and macro data to establish how fast per capita consumption and incomes grew in Ghana in the 1990s. Second, we ask how much of the rise in incomes was due to rises in the level of human capital and how much reflected underlying technical progress. Third, we assess the implications of how incomes rose for the interpretation of changes in the poverty prole. Four household surveys are used to show changes in both expenditures and incomes over the decade. The household surveys show that both consumption per capita and incomes rose by 12 per cent, a rate of 1 per cent per annum. This gure is identical to the growth rate for consumption per capita implied by the macro accounts. The average level of education of the population rose by 27 per cent over the decade which led to a rise of 3 per cent in per capita consumption. We nd, on average, no evidence for any underlying technical progress. We show that the rise in income was associated with modest falls in the head count and poverty gap measures of poverty but with virtually no change in the severity of poverty measure. The fall in the head count measure was too small to prevent the absolute number of poor people from rising. Inequality increased with the incomes of the non-agricultural self-employed, with given levels of human capital, falling both absolutely and relative to wage workers.