Auctions are a popular way to raise money for charities, but relatively little is known, either theoretically or empirically, about the properties of charity auctions. We conduct field experiments to see which sealed bid format, first price, second price or all-pay, raises the most money. Our experiment suggests that both the all-pay and second price formats are dominated by the first price auction. Our design also allows us to identify differential participation as the source of the difference between existing theory and the field.Few people appreciate the size of the philanthropic market, the amount of funding that flows through charities, and the time and resources devoted to fundraising activities each year. For example, total giving to charitable organisations in the US in 2004 amounted to nearly $250 billion (Giving USA 2005) and, according to a survey by Forbes Magazine, 200 major charities spent over $2.5 billion on fundraising activities in 2001. 1 Despite the obvious size and importance of the market for philanthropy, surprisingly little is known about the fund-raising mechanisms most likely to generate the greatest revenue for non-profit organisations.A variety of mechanisms are used to raise money for charities or to fund public goods. Secondary schools and religious congregations frequently rely on bakesales and raffles; institutes of higher education often employ student call centres and mass alumni mailings; hospitals host benefit concerts etc. (Andreoni, 2004). Interestingly, many non-profits raise revenue through auctions and, with the success of internet sites like Ebay, the popularity of charity auctions has increased. 2 The items auctioned vary to a large degree (e.g., local artwork, gift certificates for community services, weekend getaways, cars etc.), but there are relatively few auction mechanisms that are used with any regularity. One of the most common is the silent auction which corresponds closely to the standard oral ascending (or English) auction in which bids are called out sequentially. The only major difference is that participants write bids down by some pre-specified time instead of calling them out. Considering sealed bids, one may occasionally see the first price auction and to a lesser extent, the second price (or Vickrey) auction implemented. However, all-pay auctions in which participants forfeit their bids regardless of whether they win or lose are rare. We think that we know why.While the empirical literature remains thin, theory is not silent on the revenue generating properties of different charity auction mechanisms. Our immediate concern in this article is the proposition (Engers and McManus, 2002;Goeree et al., 2005)