Medical science pours its energy into developing biomarkers and drugs that enable it to predict and control patient outcomes. It spends relatively little energy, however, on inventing theories that can explain the biological processes that underly human health and disease. The current field relies heavily on data collection and on the use of statistical modeling to identify correlations in data, which can be used to predict the likelihoods of known disease trajectories. A more rational science of human health and disease, however, would rely entirely on its explanatory theories to make exact (i.e., non-probabilistic) predictions. It would spend most of its energy dreaming up imaginative theories, and then ruthlessly applying rational criticism to those theories so that it could identify flaws in them and develop them into better theories. That kind of medical science, although it would give rise to many false theories, would also give rise to astounding progress—the kind of progress that today’s mainstream medical scientists can scarcely dream of achieving. The invention of efficacious drugs would proceed, not as the dominant activity of medical science per se, but as an ancillary engineering task to be undertaken rapidly in the light of rich, deep explanatory theories of how the body works or fails to work. Discovering rich, deep explanations of human health and disease is of course difficult: but it’s possible. And yet established medical science, by spending its resources on generating data, observing correlations, and (probabilistically) predicting patient outcomes, has largely failed to achieve the kinds of explanatory breakthroughs that it needs to make awesome and dramatic progress. If the field shifts its focus, however, away from prediction toward explanation (e.g., by paying medical scientists to sit around and develop speculative theories, which will in most cases be false), then it could make that kind of progress, and not only in its immediate goal of understanding human health and disease, but also, as a consequence of that immediate goal, in its instrumental tasks of predicting and controlling patient outcomes. Indeed, by adopting this rational attitude of seeking good explanations to better understand the problems of human disease, medical science could expect, in the long run, to solve death itself.