Principles of bcnciiccnce require us to promote the good. If we believe that a plausible mom] conception will contain some such principle, we must address the issue of the demands it imposes on agents. Some writers have defended extremely demanding principles, while others have argued that only principles with limited demands are acceptable. In this paper I su ggest that we 100k at the demands 0f beneficencc in a different way; 0ur concern should not just be with the extent of (...) the demands faced by individual agents. Instead, we should consider how the demands imposed Ol'] an agent by a principle of beneiiccncc are affected by the level of compliance with the principle by others. Act—[email protected], for example, in effect requires each complying agent to shoulder her share of the demands of bcncficcnce plus as many of the shares of noncomplying agents as would be optimal. I suggest that we focus OH this feature ofconsequentialism, and not just on the very high demands it can impose Ol'] individual agents. Thus I defend the View that principles of beneiicence should not demand more of agents as expected compliance by othcr agents decreases, and formulate a principle of bencficcnce that meets this condition. This View about beneiicence and compliance is supported by a particular conception 0f beneiiccnce. Rather than as an aim we each have as individuals, benciicence could be understood as a cooperative project, where each of us aims to promote the good together with others. If s0, it would be natural t.. (shrink)