This paper replicates and refines the finding that subsidies for charitable contributions of a rebate type are less effective than matching subsidies. A survey based field experiment with health charities was conducted among a national sample representative of the Dutch population on key demographic characteristics. The greater effectiveness of matching subsidies found in laboratory experiments is replicated. Also some evidence is provided on why matches are more effective than rebates. Matches attract a larger pool of donors, in part because donors expect more people to make donations and “join in.” Matches also increase the amount contributed among the higher educated, higher income households and larger donors. Subsidies of either type do not decrease subsequent giving in a campaign for tsunami relief. The experiment could not test whether the greater effectiveness of a matching subsidy is due to a change in the donor’s attention to the benefits of a donation to the cause. This explanation should be tested in future research. The findings imply that a given budget available to subsidize charitable contributions can be used more effectively if the subsidy is framed in the form of a match than in the form of a rebate. Nonprofit organizations can use this insight in the design of fundraising campaigns. For governments the finding suggests that the effectiveness of current subsidies for charitable contributions can be enhanced by matching them rather than providing a deduction in the income tax, which works as a rebate.