It has long been known that in finite economies Classical Utilitarianism commends policies that encourage large populations. It has been known also that the stronger is the aversion to risk and inequality in the standard of living, the lower is the optimum living standard, and that the latter tends in the limit to Sidgwick's "hedonistic zero". A version of that extreme feature of the theory was subsequently named the Repugnant Conclusion (RC). Most escape routes from RC have invoked the language of "gains" and "losses", which are familiar notions in social cost-benefit analysis. Those notions have been found to lead to paradoxes involving the Non-Identity Problem. In this paper I start with Sidgwick's theory in its pristine form - the criterion for evaluating states of affairs is the sum of personal utilities - but recast it in a contemporary language: the ground of binding reason is taken to be "well-being", not "happiness", nor "agreeable consciousness". Sidgwick erred in his interpretation of the hedonistic zero, which may explain why the seeming pro-natalism inherent in his theory has been found to be repugnant by philosophers. Problems with Sidgwick's Utilitarianism lie elsewhere. An example is presented which invites an additional but relatively mild notion of person-hood into any theory that says that personal well-beings should be the sole basis for ranking states of affairs. A weak version of Agent-Relative theories is drawn from the example, which in the context of population ethics may be called Generation-Relative Utilitarianism. It has however been suggested that the theory is incoherent because it does not yield a binary relation between states of affairs. I show that the incoherence would arise only if states of affair were to be evaluated from nowhere, and that it is an essential feature of Generation-Relative Utilitarianism that the state of affairs from which other states of affairs are viewed, matters. The theory is then put to work in a model economy facing an indefinite future and a finite flow of resources. Empirical studies of Earth's life support system are then used to justify the choice of the model. I make use of contemporary global statistics to get a feel for the theory's implications for both population size and the standard of living. Population size is found to be smaller and the living standard higher than they would be under Sidgwick's Utilitarianism. Population ethics is then used to understand the nature of loss that would be suffered in the face of human extinction. Generation-Relative Utilitarianism is shown to arrive at the view that each generation is a trustee of the capital it inherits from its predecessor.