Sometimes a costly action taken by a single individual is sufficient to benefit an entire group. This should imply technical economies of scale to groups of larger size. But in a group of selfishly motivated agents, a countervailing force, the free-rider problem, may actually reduce the likelihood of provision as group size increases. Yet there are con- spicuous real-world cases where, in seeming defiance of the free-rider problem, a small minority provides a public good that benefits a large population. Examples include unpaid contributions to Wikipedia, Linux, and the bone-marrow registry. We suggest that these successful outcomes occur because a significant minority of the population is motivated, not by standard consequentialist calculations of expected benefits and costs, but by a desire to “be the one” who effects a beneficial outcome. We call persons with such motivation, Let-me-do-it types. We conduct a laboratory experiment designed toidentify such individuals, and to estimate the responsive ness of their numbers to costs and to public recognition of donors. In our experiments, we find that between 15% and 36% of subjects act as let-me-do-it types, with the proportion changing in the expected direction with costs and recognition.