Purpose: This study develops a theory of “adaptive sensemaking” in time-pressured, high-stakes medical settings to clarify the role of plausible versus accurate diagnoses in effective clinical action. To do this, the study links two influential problem-solving theories from the domain of organizational behavior: “sensemaking” and the “exploration/exploitation” dichotomy. Sensemaking theory emphasizes the role of interpretation, meaning making, and plausible stories (e.g. diagnoses) in effective action. The theory on exploring new possibilities versus exploiting known certainties concerns itself with optimizing choices and the importance of accurate perceptions of the environment. Methods: The study uses process tracing (a human factors methodology) to support in-depth qualitative analysis of how each crisis unfolds over time as well as cross-sectional quantitative comparisons of the speech and clinical actions of 39 doctors confronting the same high-fidelity, simulated operating room crisis. Results: The study finds four patterns of diagnostic sensemaking: 1) In Stalled problem solving mode, doctors make little or no progress. They neither explore a variety of plausible diagnoses nor exploit existing treatments or tests. 2) In Fixation, people generate and explore very few different plausible diagnoses, and instead heavily exploit existing treatments for a single plausible diagnosis. 3) Diagnostic Vagabonds explore many different views of the situation as they jump from plausible diagnosis to diagnosis without exploiting existing treatments or known diagnostic tests to see whether these diagnoses are accurate. 4) In Experimental problem solving, people explore a variety of plausible diagnoses and exploit multiple known treatments and tests of those diagnoses to ascertain their accuracy. Only the experimental problem solvers, who blend a moderate level of exploration and exploitation, resolve the crisis. Conclusions: This analysis provides several insights about sensemaking in crises: Exploration and exploitation are not trade-offs at two ends of a single continuum as previous research has implied, but rather, are orthogonal dimensions of problem solving. Sensemaking is also not inherently adaptive as implied in previous literature; it does not inevitably move one toward effective action. Adaptive sensemaking—sensemaking that moves toward effective clinical action— requires a blending of exploration and exploitation that allows people to generate plausible diagnosis and also test if they are accurate.