Students struggle to understand the physical basis of noncovalent interactions in the context of biochemistry. Our research goal is to develop instructional materials that take into account students' specific difficulties with biochemistry problems, principles from cognitive science, and the context of undergraduate education. From the constructivist perspective, we know pure discovery learning through problem solving alone is not beneficial; some direct instruction or guidance is necessary. Cognitive load theory suggests worked example‐problem pairs are optimal for learning due to limited working memory capacity. Alternatively, instructional design approaches in mathematics known as productive failure suggest that desirable difficulties in the initial phase of a lesson prepare students to learn from subsequent instruction. Finally, scaffolded problem solving in the form of guided inquiry is a popular instructional design where support is provided as needed and fades away as knowledge is built. Interestingly, the literature lacks a head‐to‐head comparison of these three instructional design approaches, especially in the domain of biochemistry. This study investigates the following research question: Among worked examples, productive failure, and guided inquiry, which instructional design will improve biochemistry students' problem solving about persistently difficult aspects of the physical basis of noncovalent interactions? In Spring 2018, ~100 students from an introductory biology course at a research‐intensive institution were recruited to participate in a one‐hour lesson outside of class for extra credit in their course. Participants were randomly assigned to one of three instructional conditions: (1) a worked example approach where students were guided by the instructor through two worked examples and completed two problems on their own with no guidance, (2) a productive failure approach where students explored a set of two problems followed by instructor‐led direct instruction through two worked examples, and (3) a guided inquiry approach where students worked through four problems with guidance from an instructor and peer learning assistants. After completing the instructional phase, participants completed an assessment that included near and far transfer problems. Prior knowledge was equivalent in all instructional groups, and each instructional approach yielded normalized knowledge gains. Comparisons of near transfer problem‐solving performance revealed a statistically significant instructional effect. Specifically, participants in the worked example and productive failure instructional groups outperformed participants in the guided inquiry group. In future work we will further explore the impact of instruction on far transfer problem‐solving as well as the extent to which variables like spatial ability influence the impact of instruction.Support or Funding InformationARCS FoundationThis abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.