Low and variable farm income has been a main rationale for heavy government intervention in agricultural markets and income transfers to farmers whether in Europe in response to disruptive agricultural imports and low world prices at the end of the 19th century or in the US in response to the Great Depression. While the future of the Common Agricultural Policy (CAP) is again discussed and new directions are examined, it is fundamental to know to what extent low and variable farm income is still a problem in contemporary European agriculture and a valid rationale for designing the new CAP. In this context, this paper first examines the income level and distribution of farm households compared to those of non-farm households for a selection of OECD countries. Second, the paper econometrically investigates whether explanations for low farm income given in the literature apply to the selected OECD countries for the 1980-2000 period. Third, the paper concludes with some policy implications. Both the descriptive and econometric analyses use the microeconomic dataset from the Luxembourg Income Study (LIS). This dataset contains socio-demographic, income and expenditure data that are collected at the household level through household-based budget surveys. These data are recorded in the LIS dataset in a harmonized way for the 30 countries that currently participate in the LIS. Average income levels as well as indicators of poverty and inequality are calculated for farm and non-farm households for the OECD countries that have at least three waves of data in the LIS dataset with a minimum of 30 identified farm households surveyed in each wave1. Three sets of explanations for low farm household income drawn from the literature review of Gardner (1992) are successively investigated: (i) the commodity market conditions, (ii) the earning disequilibrium between sectors, and (iii) the compensating differential for skill differences and non-pecuniary aspects2. Preliminary results confirm that in most of the 12 selected OECD countries the average farm household income is greater than the average non-farm household income. Lower average farm household income tends to occur sporadically for some years in only six of the 12 selected OECD countries. In five of the nine selected European countries, the average farm household incomes clearly tend to improve compared to the average non-farm household incomes during the 1985-95 period. They are well above the average household incomes. The incidence of poverty tends to be less severe among farm households than nonfarm households except for two European countries. In contrast, the intensity of poverty tends to be more severe among farm households than non-farm households in most countries. This implies that in general there are relatively fewer poor farm households compared to non-farm households but the severity of their poverty is stronger. In addition, the income distribution is more equal among farm households than non-farm households in all countries. In the final version of the paper, it is expected that each set of explanations for low farm household income would play a role in understanding the evolution of farm household incomes across countries. Depending on whether one set of explanations tends to dominate the others in one particular group of countries, public interventions can be designed and emphasized in a new CAP for improving and stabilising incomes of farm households.