Background: Caffeine citrate is the methyl-xanthine of choice used in controlling apnea of prematurity (AOP). Caffeine central effect is mediated via non-selective (A1) and selective (A2a) adenosine receptors antagonism. Variability in caffeine response had been frequently noticed in AOP, suggesting underlying genetic predisposition. Aim of the study: We evaluated the role of adenosine receptor A1 [ADORA1] and adenosine receptor A2a [ADORA2a] gene polymorphisms in the variability of caffeine response among Egyptian preemies with AOP. Patient and methods: In this case-control study, 43 preterm neonates with AOP were eligible as cases and 43 preterm babies free from apnea were taken as controls. Preterm neonates with AOP were further divided according to response to caffeine treatment into caffeine responder (n = 18) and caffeine non-responders (n = 25). ADORA1 [716 T > G] and ADORA2a [1976C > T] gene polymorphisms were genotyped by mean of PCR-based RFLP-assay. Results: There were significant increase in frequency distribution of ADORA2a [1976C > T] CT (62.7% vs 23.3%), TT (14% vs 4.7%) genotypes and T allele (34.3% vs 16.3%) in cases compared to controls with significant increased risk of AOP development with OR (95%CI); P-value of 8.37(3.03–23.1), P = 0.000; 9.3(1.61–53.61), P = 0.005 and 4.27(2.09–8.70), P = 0.000 respectively. Further, caffeine non-responders were associated with significant increase frequency of ADORA2a CT (80% vs 38.9%) and TT (16% vs 11.1%) genotypes and T allele (56% vs 30.6%) with OR (95%CI) and P-value of 21.38 (2.31–197.8), P = 0.001; 18 (1.24–260.9); P = 0.005 and 2.89(1.17–7.13), P = 0.019 respectively, when compared to caffeine responders. Patients with AOP who had ADORA2a CT and TT genotypes were associated with significant increase in duration of hospital stay and poor outcome. Genotype distribution frequency of studied polymorphisms did not deviate from Hardy Weinberg (HW) equilibrium among controls. Conclusion: ADORA2a [1976C > T] polymorphism has a significant role in AOP development and variation in caffeine response among preterm babies