Many studies have shed light on rater criteria for assessing the performance of language skills (e.g., Eckes, 2005). However, the interface between rater assessment and interlanguage pragmatics (ILP) has remained largely unnoticed. To address this interface, this study explored the ratings native (NES) and nonnative English speaking (NNES) teachers assigned to second language (L2) refusal production and the criteria they applied in their ratings. To this end, 50 NES and 50 NNES teachers participated in rating L2 refusal production of EFL learners that included responses to a 6-item written discourse completion task. The data were analyzed qualitatively and quantitatively. Qualitative analysis showed that native teachers applied 11 criteria and nonnative teachers applied 6 criteria in their pragmatic ratings. Reasoning/explanation was the leading criterion in teacher assessment among native raters, whereas politeness was the main criterion for nonnative ratings. Quantitative analysis documented variation in the frequency of drawing on rating criteria and significant differences in ratings, with NNES teachers being more lenient and divergent in their ratings. The results suggest there is a gap between NES and NNES teachers in terms of rating criteria, strictness, and convergence in rating.