Apnoea of prematurity (AOP) is among the most common diagnoses in the neonatal intensive care unit. Caffeine treatment is a preferred treatment choice. However, neonatal caffeine therapy results in significant intersubject variability. This study aimed to determine the effects of plasma caffeine levels based on standard dose and genetic variability on clinical response to caffeine citrate in Chinese preterm infants. This single-center and retrospective study examined data from 112 preterm infants (< 35weeks gestational age) between July 2017 and July 2018. Subjects were divided into apnoea-free (n = 48) and apnoeic (n = 64) groups, and their clinical outcomes were summarized. Liquid chromatography-tandem mass spectrometry was used to measure levels of caffeine and its primary metabolites. Eighty-eight single-nucleotide polymorphisms were chosen for genotyping by a MassARRAY system. Preterm infants in the apnoea-free group were associated with a reduction in the incidence of bronchopulmonary dysplasia and a reduced requirement for patent ductus arteriosus ligation. No significant association was observed between plasma-trough-concentration-to-dose (C0/D) ratio and birth weight, gestational age, or postnatal age in either group. Polymorphisms in CYP1A2 and aryl hydrocarbon receptor (AHR) genes did not affect plasma caffeine levels. Polymorphisms in adenosine receptor genes ADORA1 (rs10920568 and rs12744240), ADORA2A (rs34923252 and rs5996696), and ADORA3 (rs10776727 and rs2298191), especially in AHR (rs4410790) and adenosine deaminase (rs521704), play critical roles in the interindividual response to caffeine therapy. Genetic polymorphisms in caffeine's target receptors, but not the exposure levels based on the standard dosing, were associated with variable responses to caffeine therapy in preterm neonates. Future studies are needed to uncover how these genetic variants affect responses to caffeine therapy in this patient population.