BackgroundNurses have a crucial role in managing, educating, and caring for diabetic patients. However, their knowledge should be regularly assessed to avoid preventable complications and reduce costs. Therefore, the present study assessed the perceived and actual knowledge about diabetes among nurses in Hail province of Saudi Arabia and investigated predictors of such knowledge.MethodsA cross-sectional study was conducted among 325 conveniently sampled nurses from all public hospitals and primary healthcare centres in Hail province from September to December 2022. A pre-designed questionnaire was used to collect demographic and practice-related characteristics of the nurses. In addition, data on nurses’ perceived and actual knowledge about diabetes were collected using the Diabetes Self-Report Tool (DSRT) and Diabetes Basic Knowledge Tool (DBKT) self-report questionnaires, respectively. The mean knowledge scores for demographic and practice-related variables were compared using the independent-samples t-test and one-way analysis of variance. Multiple linear regression was used to identify significant predictors of perceived and actual knowledge. The correlation between perceived and actual knowledge was investigated using Pearson’s correlation coefficient. A P-value ˂0.05 was considered statistically significant.ResultsBased on a highest maximum score of 60 using the DSRT, the mean score of perceived knowledge was 38.4 ± 12.0, corresponding to a percentage mean score of 64%. On the other hand, based on a highest maximum score of 49 using the DBKT, the mean score of actual knowledge was 23.2 ± 9.6, corresponding to a percentage mean score of 47.3% of correct responses. Being Indian, having a diploma or a bachelor’s degree, and having a poor or fair self-perception of competence in diabetes care were predictors of lower perceived knowledge scores, whereas having no access to diabetes guidelines was a predictor of higher scores. However, being non-Saudi and having experience of at least 16 years were predictors of higher actual knowledge scores. The correlation between actual and perceived knowledge about diabetes was negligible and statistically non-significant (r = 0.011, P = 0.055).ConclusionNurses affiliated with public health facilities in Hail province lack adequate knowledge about diabetes, with no correlation between what is perceived to be known and what is actually known. Indian citizenship, having a diploma or bachelor’s degree, not having access to diabetes guidelines, not attending courses/workshops, and having a poor or fair self-perception of competence in diabetes care can significantly predict nurses’ perceived knowledge. However, being non-Saudi (Filipino or Indian) and having at least 16 years of experience can significantly predict their actual knowledge of diabetes.