Background: Research has shown that patients' and practitioners' cultural orientations affect communication behaviors and interpretations in cross-cultural patient-practitioner interactions. Little is known about the effect of cultural orientations on assessment of communication behaviors in cross-cultural educational settings. The purpose of this study is to explore cultural orientation as a potential source of assessor idiosyncrasy or between-assessor variability in assessment of communication skills. More specifically, we explored if and how (expert) assessors' valuing of communication behaviours aligned with their cultural orientations (power-distance, masculinityfemininity, uncertainty avoidance, and individualism-collectivism). Methods: Twenty-five pharmacist-assessors watched 3 videotaped scenarios (patient-pharmacist interactions) and ranked each on a 5-point global rating scale. Videotaped scenarios demonstrated combinations of well-portrayed and borderline examples of instrumental and affective communication behaviours. We used stimulated recall and verbal protocol analysis to investigate assessors' interpretations and evaluations of communication behaviours. Uttered assessments of communication behaviours were coded as instrumental (task-oriented) or affective (socioemotional) and either positive or negative. Cultural orientations were measured using the Individual Cultural Values Scale. Correlations between cultural orientations and global scores, and frequencies of positive, negative, and total utterances of instrumental and affective behaviours were determined. Results: Correlations were found to be scenario specific. In videos with poor or good performance, no differences were found across cultural orientations. When borderline performance was demonstrated, high power-distance and masculinity were significantly associated with higher global ratings (r = .445, and .537 respectively, p < 0.05) as well as with fewer negative utterances regarding instrumental (task focused) behaviours (r = -.533 and - .529, respectively). Higher masculinity scores were furthermore associated with positive utterances of affective (socioemotional) behaviours (r = .441). Conclusions: Our findings thus confirm cultural orientation as a source of assessor idiosyncrasy and meaningful variations in interpretation of communication behaviours. Interestingly, expert assessors generally agreed on scenarios of good or poor performances but borderline performance was influenced by cultural orientation. Contrary to current practices of assessor and assessment instrument standardization, findings support the use of multiple assessors for patient-practitioner interactions and development of qualitative assessment tools to capture these varying, yet valid, interpretations of performance. [ABSTRACT FROM AUTHOR]