Background and purpose: Hypoglycemia is an important problem in newborns which requires early diagnosis and treatment. Laboratory testing of blood glucose level and glucometer readings are used to detect hypoglycemia but sometimes the results are given late. This study aimed at evaluating the accuracy of glucometer in detection of hypoglycemia in newborns at neonate intensive care unit (NICU), 2014-2015. Materials and methods: A diagnostic study was performed in which 408 blood samples (glucometer capillary and venous derived samples) were obtained from 204 neonates in NICU within two hours after birth. Blood glucose concentrations were measured by glucose oxidase and photometric analysis. The mean values for blood glucose level, correlation coefficient, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were determined. Data analysis was performed in SPSS V.19. Results: The mean neonatal weight was 2915 ± 840 gr. In our study, 92 neonates were females and 112 were males. In laboratory testing the prevalence of hypoglycemia was found to be 21.6% (CI 95% 15.9-27.3%). Mean values of blood glucose level (86.4 ± 60.9 mg/dl) with glucometer was significantly higher than that of laboratory level (77.8 ± 69.1 mg/dl), (P<0.001). At cutoff point of 40 mg/dL blood glucose level the diagnostic indices of glucometer were as follows: sensitivity 95.6%, specificity 45.5%, PPV 86.4%, and NPV 74.1%. According to ROC curve, 49 mg/dl was obtained as the glucometer cutoff point and the prevalence of hypoglycemia was found to be 15.2% (CI 95% 10.2-20.2%). Conclusion: Glucometer was found to have appropriate sensitivity and NPV in detecting neonatal hypoglycemia, therefore, it is recommended in screening of neonatal hypoglycemia without any side effects.