Small renal masses (SRMs) have been shown to have low malignant potential. Active surveillance (AS), typically characterized by regular follow-up and delayed nephrectomy if necessary, is recommended as an option for frail patients with SRMs. Nevertheless, the impact of tumor size on survival in T1a RCC patients undergoing delayed nephrectomy for SRMs remains unclear. Patients diagnosed with non-metastatic T1a RCC who underwent nephrectomy were identified from the Surveillance, Epidemiology, and End Results (SEER) database and divided into immediate (< 6 months) and delayed nephrectomy (≥ 6 months) groups based on the duration from diagnosis to nephrectomy. After propensity score matching (PSM), overall survival (OS) and cancer-specific survival (CSS) were estimated by K-M curves and compared with log-rank test. A total of 27,502 patients were enrolled, of whom 26,915 (97.9%) received immediate nephrectomy and 587 (2.1%) received delayed nephrectomy. After PSM, 1174 patients who underwent immediate nephrectomy and 587 patients who underwent delayed nephrectomy were included. With a median delay of 7 months, delayed nephrectomy resulted in non-inferior OS for RCC tumors sized 0.1–2.0 cm (HR = 1.12, p = 0.636). However, for RCC tumors sized 2.1–3.0 cm (HR = 1.60, p = 0.008) and 3.1–4.0 cm (HR = 1.89, p < 0.001), delayed nephrectomy showed inferior OS compared to immediate nephrectomy. Delayed nephrectomy did not result in significantly worse CSS than immediate nephrectomy in all tumor size subgroups (all p > 0.05), however this may be due to sample size limiting statistical power. Based on the SEER database, we found that with a median delay of 7 months, 2 cm may be an appropriate cut-off point of delayed nephrectomy for patients diagnosed with non-metastatic T1a RCC.