Using two linked simulation models, we examine the public expenditure costs and distributional effects of potential reforms to long-term care funding in the UK. Changes to the means tests for user contributions to care costs are compared with options for the abolition of these means tests (‘free’ personal care). The latter generally cost more than the former and benefit higher income groups more than those on lower incomes (measuring income in relation to the age-specific income distribution). Reforms to the means tests target benefits towards those on lower incomes. However, the highest income group are net losers if free personal care is financed by a higher tax rate on higher incomes and the effect on the whole population considered.