Introduction Clinical practice variability is characterized by two or more expert clinicians who make different treatment decisions despite encountering a similar case. Thrombosis Medicine is an example of a specialty that has a large evidence base to support decision-making and where practice variability is common. Our objective was to explore how residents experience and interpret inter-supervisor clinical practice variability in Thrombosis Medicine, and how these variations influence learning. Methods Seventeen senior residents in Internal Medicine, Hematology or Thrombosis Medicine (Postgraduate Year [PGY] 3-6) participated in semi-structured interviews. Data collection and analysis occurred iteratively and concurrently in a manner consistent with constructive grounded theory methodology. Variation theory was used as a sensitizing concept. A central tenet of this theory is that learning occurs by experiencing three sequential patterns of variation: contrast, generalization, and fusion. All transcripts were coded in duplicate by paired members of the study team. Participants were recruited purposively with regard to their field of training until thematic saturation was reached. The primary investigator maintained an audit trail and reflexivity journal. Member checking was performed. Dedoose (version 8.0.35) was used for all analyses. The Hamilton Integrated Research Ethics Board reviewed the study. Results Clinical practice variability was common and generally viewed positively. Residents attributed clinical practice variability to supervisor personality differences, inter-institutional differences, availability and interpretation of evidence, differences in patient preferences and characteristics, and their own participation in the decision-making process. Inherent differences between supervisors, including personality traits, risk tolerance, and prior educational and clinical experiences were a major contributor. Clinical practice variability was felt to be most common when the clinical problem had scant or conflicting evidence to guide practice, for example, in the treatment of subsegmental pulmonary embolism or splanchnic vein thrombosis. Supervisors and learners often took patient preferences into account when making a decision, especially when the net clinical benefit of anticoagulation was uncertain. Supervisor-learner discussions were critical to learning from clinical practice variability. Through these interactions, residents asked clarifying questions, and supervisors shared tacit knowledge, elaborated on their prior experiences, compared and contrasted similar cases, and articulated their clinical reasoning. Residents endorsed the view that a certain baseline level of knowledge is required to appreciate the nuances that are uncovered by clinical practice variability. Junior residents (PGY 2-4) were often assessment-focused and made note of practice patterns between their supervisors in order to predict their supervisors' preferences for future cases. More experienced residents (PGY 5-6), especially those pursuing advanced training in Thrombosis Medicine, were more likely to view clinical practice variability as a means to improve their own decision-making skill in preparation for independent practice. Consistent with variation theory, clinical practice variability helped residents to discern critical aspects that influenced decision-making (contrast), group similar cases together so that the appropriate evidence could be applied (generalization), and weigh multiple variables to make a treatment decision (fusion). Observing practice variability was more helpful for senior learners and less so for junior learners as inexperience and knowledge gaps made it difficult to reconcile differences. Conclusion Residents had a generally positive perception of clinical practice variability in the workplace. Practice variability helped residents discern critical aspects, group similar patients together and practice individualized medicine. Reflection, literature review, and case discussions were felt to maximize learning from practice variability. Clinical supervisors and curriculum designers can promote learning from practice variability through first and second order scaffolding, modeling, coaching, and thinking aloud.