Purpose. To determine what learners and teachers value most in ambulatory learning encounters and whether the choices of the two groups are in concordance. Method. In 1996, the authors surveyed learners and teachers at a walk-in clinic immediately after each of 103 consecutive learning encounters. The participants answered two open-ended questions: (1) What was the one most valuable aspect of this learning encounter? and (2) List one thing that would make it better. Using qualitative analysis methods, two raters categorized the responses; their agreement was substantial (κ = .75). Results. Half the responses fit five categories: diagnosis (15%), general management (14%), physical examination skills (9%), patient selection (6%), and time issues (6%). The participants most valued general exposure to diagnosis (29%) and general management issues (29%). Regarding their suggestions for improving the encounters, 33% cited structural issues (often 'more time'), while 43% had no suggestions ('nothing' or 'fine as is'). Substantial concordance existed between the rankings by category of the teachers' and learners' comments, but (1) learners were more likely to rate the educational value of the encounter excellent or very good (64% vs 47%, p < .01); (2) teachers were more likely than medical students (but not interns) to place the most educational value on the physical examination (30% vs 4%, p = .001); and (3) agreement on what was most valuable in any individual encounter was poor (κ = .03). Conclusions. In evaluating ambulatory educational encounters, learners and teachers placed highest value on general exposure to diagnosis and disease management; while the most commonly recommended changes related to structural issues, particularly inadequate time. While learners and teachers agreed in general on the relative ranking of teaching activities, they often found different things to be educationally salient in a particular encounter.