Purpose The most appropriate time of primary tumor radiotherapy in non-small cell lung cancer(NSCLC) with EGFR-TKIs remains unclear. The aim of this study was to investigate the effect of the time factor of primary tumor radiotherapy on long-term overall survival(OS)and provide a theoretical basis for further clinical research. Patients and methods In total, 238 patients with EGFR-TKIs and OS >= 12 months were statistically analysed. Patients were grouped: the D group without primary tumor radiotherapy and the R group with it. The R group were divided into three groups according to the interval between the start of EGFR-TKIs and the start of primary tumor radiotherapy: R0 - 30(<30 days), R30 - PD(>= 30 days and disease stable), and R-PD(radiotherapy after disease progression). The Kaplan-Meier method and log-rank test were used for survival analyses. Exploratory landmark analyses were investigated. Results The OS rates at 1, 2, 3, 5 years for the R group and D group were 96.8%, 62.9%, 38.3%, 17.1%, and 95.6%, 37.7%, 21.8%, 2.9%, respectively; the corresponding MST was 29 months(95% CI: 24.3-33.7) for the R group and 22 months(95% CI: 20.4-23.6) for the D group (chi(2) = 13.480, p<0.001). Multivariate analysis revealed that primary tumor radiotherapy was independent predictors of prolonged OS. Among the four groups, The R30 - PD appeared to have the best OS (D, chi(2) = 19.307, p<0.001;R0 - 30, chi(2) = 11.687, p = 0.01; R-PD, chi(2) = 4.086, p = 0.043). Landmark analyses(22 months) showed the R30 - PD group had a significant long-term OS. The incidence of radiation pneumonitis >= grade 2 was17.3%(n = 19)and radiation esophagitis >= grade 2 was observed in 32 patients(29.1%). Conclusions Our results showed that primary tumour radiotherapy may prolong long-term OS with acceptable toxicities. Appropriate delay(R30 - PD)of primary tumour radiotherapy may be the best choice. Premature radiotherapy(R0 - 30) and radiotherapy after disease progression (R-PD)may not be reasonable for long-term OS.