Materials and Methods: This was a cross-sectional study implemented on all midwives (202) occupied in Tabriz health care centers, Iran, using socio-demographic, knowledge and practice questionnaires. The instrument validity was determined based on content validity index (CVI) and CVR methods while its reliability was determined from 2 aspects of repeatability (ICC) and internal consistency (Cronbach alpha coefficient). The socio-demographic and knowledge data questionnaires were completed by midwives while observational checklist of practice was completed by researcher. The data were statistically analyzed using descriptive and inferential methods.

Results: Mean (SD) score of midwives? knowledge was 73.21 (± 11.83) and mean (SD) score of midwives? practice was 80.31 (±12.25). According to Pearson test, the direct relation between knowledge and practice was not significant (P = 0.133, r= 0.238). According to multivariate linear regression analysis, age, educational level, employment status and work experience were defined as predictors of midwives? knowledge and age, employment status and job responsibilities as predictors of midwives? practice.

Conclusion: With regard to the direct correlation between knowledge and practice, improvement of midwives? knowledge and so paying attention to the predictors of practice such as employment status and job responsibilities in order to promote midwife?s motivation and practice seems to be essential to improve the quality of preconception care.