How can altruistic organisms that sacrifice their fitness to help their competitors survive natural selection? This book attempts to resolve this puzzle by showing that it stems from the questionable assumption that an altruistic agent necessarily performs selectively disadvantaged acts. Altruism can be selectively advantaged in social settings in which reputation is a criterion for selecting partners for transactions The book develops an analytically precise conception of altruism and a new version of the Prisoners Dilemma which allows for opting-out of transactions. The analysis develops criteria for the frequency and magnitude of altruistic acts that enable altruistic agents to invade and persist in populations of cooperators and egoists. Also investigated is the implicit assumption of all models (including the ones in this book) that analyze the evolution of pro-social behaviors: there are groups of individuals with sufficient cohesion to enable a large number of interactions. The book presents a computer simulation based upon the foraging behavior of the great apes that studies the conditions for formation of such coalitions from independent, egoistic organisms.