Background: Although caffeine is an effective drug in premature babies, there is still no consensus regarding its effective dose. This interventional study aimed at the effect of caffeine with a low maintenance dose compared to a high dose on the respiratory function of premature infants of ≤ 34 weeks. Methods: In this clinical trial, 60 premature babies with a gestational age ≤34 weeks diagnosed with respiratory distress syndrome were included in the study. Patients randomly and equally received an initial dose of 20mg/kg and then a maintenance dose of 10mg/kg and 5mg/kg of caffeine. Extubation failure rate, CPAP duration, hospitalization duration, adverse effects and outcome of patients were compared in two groups. Results: Both groups had no statistically significant differences in terms of basic variables. The frequency of failure in extubation (36.7 vs. 13.3%), the duration of mechanical ventilation (6.4 ± 2.8 vs. 3.5 ± 0.5 days), the duration of NCPAP requirement (8.2 ± 6.3 vs. 4.3 ± 2.7 days) and the number of apneas (2.7 ± 2.4 vs. 1.9±1.6) in the group with higher dose was found to be lower as compared to other group, where a statistically significant difference was found. There was no statistically significant difference between the two groups in terms of side effects. Conclusion: Our data has shown the effectiveness of a maintenance dose of 10 mg caffeine compared to 5 mg caffeine, which can be associated with a better short-term outcome in premature babies.