Background Prematurity is a well-known risk factor for asthma. Biomarkers from sputum or serum were determined to have association with childhood asthma but no existing study has focused on salivary cytokines nor on prematurity. Thus, our study aimed to determine the relationship between childhood asthma and salivary cytokines in the early life of prematurity. Methods Preterm neonates from August 2012 to May 2017 were recruited and those with bacterial infection within seven days of life, maternal sepsis, and maternal clinical chorioamnionitis were excluded. Their gestational age, birth body weight, comorbidities, intubation duration, and infectious times during hospitalization were documented. The salivary cytokines on the first (D1) and seventh (D7) day of life were detected by MILLPLEX® MAP Human multiplex assay. They were followed up and divided into asthma and non-asthma groups. Kaplan-Meier or Cox-regression were used for analyzing the data. Results A total of 125 preterm infants were enrolled with 21 in the asthma group and 104 in the non-asthma group. Patients in the asthma group had longer duration of intubation (p < 0.001) and higher re-hospitalization rate for respiratory tract infection (p < 0.001), especially for respiratory syncytial virus bronchiolitis (p = 0.008). For salivary cytokines, the levels of D1 interleukin (IL)-8 were significantly higher in the asthma group (p = 0.04). Conclusions Salivary IL-8 at one day of life was associated with the development of asthma in childhood of premature neonates. Therefore, we can intervene early on these high-risk premature infants to prevent the occurrence of asthma.