In patients with epidermal growth factor receptor (EGFR)-mutant non-small-cell lung cancer (NSCLC) with brain metastases, it remains controversial whether the use of EGFR-tyrosine kinase inhibitor (TKI) alone without radiotherapy (RT) is an optimal approach. Here, we investigated the clinical outcomes according to the use of upfront RT as well as the subsequent therapy following intracranial progression. This single-centre retrospective study included a total of 173 patients who were treated with EGFR-TKI alone (TKI alone group) or with upfront whole-brain RT (WBRT) or stereotactic radiosurgery (SRS) followed by EGFR-TKI (RT plus TKI group). Clinical outcomes according to initial and subsequent therapies following intracranial progression were analysed. There was no significant difference in OS according to the use of upfront RT (TKI alone group, 24.5 months vs. WBRT group, 20.0 months vs. SRS group, 17.8 months; P = 0.186). Intracranial progression was found in 35 (32.7%) of 107 patients in the TKI alone group. Among them, 19 patients who received salvage RT had the better prognosis than others [median overall survival (OS); 28.6 vs. 11.2 months; P = 0.041]. In the RT plus TKI group, 12 (18.1%) of the 66 patients experienced intracranial progression and 3 of them received salvage RT (median OS; 37.4 vs. 20.0 months; P = 0.044). In multivariate analysis, upfront WBRT was associated with trends towards a lower probability of intracranial progression, whereas upfront SRS was found to be an independent risk factor for poor OS. In conclusion, using EGFR-TKI alone for brain metastasis in EGFR-mutant lung cancer patients showed outcomes comparable to those using upfront RT followed by EGFR-TKI. Patients who could not receive salvage RT following intracranial progression had the worst survival regardless of the type of initial treatment.