Research in science, technology, engineering, and mathematics education supports a shift from traditional lecturing to evidence-based instruction in college courses, yet it is unknown whether particular evidence-based pedagogies are more effective than others for learning outcomes like problem solving. Research supports three distinct pedagogies: worked examples plus practice, productive failure, and guided inquiry. These approaches vary in the nature and timing of guidance, all while engaging the learner in problem solving. Educational psychologists debate their relative effectiveness, but the approaches have not been directly compared. In this study, we investigated the impact of worked examples plus practice, productive failure, and two forms of guided inquiry (unscaffolded and scaffolded guidance) on student learning of a foundational concept in biochemistry. We compared all four pedagogies for basic knowledge performance and near-transfer problem solving, and productive failure and scaffolded guidance for far-transfer problem solving. We showed that (1) the four pedagogies did not differentially impact basic knowledge performance; (2) worked examples plus practice, productive failure, and scaffolded guidance led to greater near-transfer performance compared with unscaffolded guidance; and (3) productive failure and scaffolded guidance did not differentially impact far-transfer performance. These findings offer insights for researchers and college instructors.