OBJECTIVES: To report on the reliability and validity of an instrument used among diabetic patients who smoke. The instrument that was designed and pretested is a 41-item questionnaire to evaluate knowledge on diabetes, physician practice, and their attitudes towards a smoking cessation intervention when applied by their physicians. METHODS: We developed the questionnaire based on information drawn from the literature related to diabetes care and management and smoking negative effect. Questions were adapted and modified from other relevant literature. To ensure face, content and construct validity, the questionnaire was reviewed by six pharmacists, five of whom were faculty members with experience and expertise in survey instruments development and research. Modifications suggested included, but not limited to adding multiple choice responses to some of the knowledge and practice items, moving some of the knowledge items to the practice section, adding questions to the knowledge domain, and lastly putting the attitude questions in a table format which was easier to be understood. The questionnaire was modified accordingly and sent for a second review. The questionnaire was readable by respondents with at least grade 6 education level. The reconciled and modified version was sent to be translated to Malay language, which is the national language of Malaysia. The translation was validated using the standard forward and backward method. RESULTS: The questionnaire was pretested on a convenient sample of 85 patients. Reliability analysis of the questionnaire using Cronbach’s alpha showed an internal consistency reliability of 0.9 for the attitude domain. The reliability of four knowledge items was measured by the split-half reliability method; the Spearman-Brown and Guttman Split-Half Coefficients were 0.6;  and 0.56, respectively.  CONCLUSIONS: The questionnaire was valid and reliable to evaluate diabetic knowledge, physician practice, and attitude towards the smoking cessation intervention when applied by their physicians.