Background: Recurrence is a key outcome to evaluate the treatment effect of differentiated thyroid carcinoma (DTC). However, no consistent definition of recurrence is available in current literature or international guidelines. Therefore, the primary aim of this systematic review was to delineate the definitions of recurrence of DTC, categorized by total thyroidectomy with radioactive iodine ablation (RAI), total thyroidectomy without RAI and lobectomy, to assess if there is a generally accepted definition among these categories. Methods: This study adhered to the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement. In December 2023, a systematic literature search in MEDLINE and EMBASE was performed for studies reporting on the recurrence of DTC, from January 2018 to December 2023. Studies that did not provide a definition were excluded. Primary outcome was the definition of recurrence of DTC. Secondary outcome was whether studies differentiated between recurrence and persistent disease. Two independent investigators screened the titles and abstracts, followed by full-text assessment and data extraction. The study protocol was registered in PROSPERO, CRD42021291753. Results: In total, 1450 studies were identified. Seventy studies met the inclusion criteria, including 69 retrospective studies and 1 randomised controlled trial (RCT). Median number of patients in the included studies was 438 (range 25-2297). In total, 17 studies (24.3%) reported on lobectomy, 4 studies (5.7%) on total thyroidectomy without RAI, and 49 studies (70.0%) with RAI. All studies defined recurrence using one or a combination of four diagnostic modalities cytology/pathology, imaging studies, thyroglobulin (-antibodies), and a predetermined minimum tumor-free time span. The most common definition of recurrence following lobectomy was cytology/pathology-proven recurrence (47.1% of this subgroup), following total thyroidectomy with RAI was cytology/pathology-proven recurrence and/or anomalies detected on imaging studies (22.4% of this subgroup). No consistent definition was found following total thyroidectomy without RAI. Nine studies (12.9%) differentiated between recurrence and persistent disease. Conclusion: Our main finding is that there is no universally accepted definition for recurrence of DTC in the current studies across any of the treatment categories. The findings of this study will provide the basis for a future, international Delphi-based proposal to establish a universally accepted definition of recurrence of DTC. A uniform definition could facilitate global discussion and enhance the assessment of treatment outcomes regarding recurrence of DTC.