High-dosage tutoring is receiving a lot of buzz as a promising tool to address learning loss in the wake of the COVID-19 pandemic. But unlike vaccines, successful tutoring programs are challenging to scale with fidelity. In this paper, long-time educators Michael Goldstein and Bowen Paulle recommend: (1) Evaluating tutoring programs and measuring their results; (2) Abandoning those that don't work; and (3) Conditionally scaling those that do work in small settings, all while focusing on desired program outcomes, not inputs. To be sure, high-dosage tutoring helps many children. But effectively scaling up small programs is no easy feat. Context and quality matter. If policymakers make well-informed choices and do more than just throw money at these programs, high-dosage tutoring can be grown to benefit many more students than it already does.