Low replication rates and ill-informed selection of replication targets can harm clinical practice. We take a pragmatic approach to suggest which studies to replicate in clinical psychology. We propose a 2-step process entailing a quantitative assessment of evidence strength using Bayes factors, and a qualitative assessment concerning theory and methodology. We provide proof of concept on a series of published clinical studies. We included 75 studies with 94 individual effects. Step 1 yielded 42 effects (45%) with Bayes Factors suggesting ambiguous evidence or absence of an effect. These 42 effects were qualitatively assessed by 2 raters. We illustrate their decision process and discuss advantages and disadvantages of the proposed steps.