It has been common in contemporary logic and philosophy of logic to identify the validity of an inference with its conclusion being a (logical) consequence of its premisses. This identifcation pays attention to at most a necessary condition for an inference being acceptable in a deductive argument or proof. An inference is not acceptable unless the conclusion becomes evident because of being supported by the premisses. Can we defne this condition in a stringent way so that we get a concept of valid inference allowing us to characterize a proof or valid deductive argument as a chain of valid inferences? This is the main question that I shall be concerned with in this essay. By the validity of an inference I understand henceforth a concept of that kind, which I shall strive to explicate here.