Introduction Although positron‐emission tomography (PET) plays an integral role in cervix cancer diagnosis, there are limited data on PET‐based image‐guided brachytherapy (IGBT). We aimed to report the long‐term outcomes of PET‐based IGBT. Methods We reviewed 151 patients treated with definitive radiotherapy (RT), including PET‐based IGBT between 2009 and 2018. After median 45 Gy of external beam RT with the four‐field technique, a median 24 Gy of high‐dose‐rate iridium‐192 IGBT was delivered in six fractions with Fletcher‐Suit tandem and ovoids. All patients underwent 18F‐fluorodeoxyglucose‐PET/computed tomography planning with a brachytherapy applicator. Multivariable analysis of local control (LC) was performed using Cox regression analysis. Results The median high‐risk clinical target volume (HRCTV) and HRCTV D90% were 51.8 (interquartile range IQR 35.9–79.4) cm3 and 77.7 (IQR 74.7–81.2) Gy, respectively. With a median follow‐up of 57 (IQR 24.3–81.4) months, the 5‐year LC rate was 89.2%. HRCTV ≥72 cm3 was associated with inferior LC (hazard ratio, 3.72, p = .017) after multivariable analysis: the 5‐year LC rates were 94.0% and 77.9% for HRCTVs ≥72 and < 72 cm3, respectively (p = .002). The impact of HRCTV D90% ≥70 Gy on LC was significant in patients with an HRCTV ≥72 cm3 compared to that in those with HRCTV < 72 cm3. Patients with adeno/adenosquamous carcinoma demonstrated inferior LC in both groups. There were 13 (8.6%) and 11 (7.3%) patients with acute and late severe toxicities after RT. Conclusion PET‐based IGBT leads to favorable LC, and HRCTV ≥72 cm3 requires further dose escalation to improve outcomes. Development of a risk prediction model to reliably identify patients undergoing ALL treatment at a higher risk of life threatening sepsis based on analysis of 2068 fever episodes in 377 patients treated on the consecutive Malaysia‐Singapore ALL clinical trials.