BACKGROUND Critical thinking (CT) is a composite of skills linked to problem-based learning (PBL). PURPOSES This study has 3 purposes: (a) to determine if PBL discourse could be coded for CT, (b) to demonstrate reliable coding, and (c) to determine whether a CT ratio would provide a valid measure to compare 2 PBL groups. METHODS Using prior research, we refined the code for a content analysis of PBL transcripts. Raters coded 6 hr of transcripts and computed CT ratios for each of the 5 CT stages. Average interrater agreement was 85.5%. CT ratios appeared to differ between 2 PBL groups delivered in 2 modalities. RESULTS PBL discourse could be coded following a CT framework. Independent raters reliably applied the code, and the resulting CT ratios detected tenable differences. CONCLUSIONS This approach could provide useful information about the effect of case modality.