Introduction Thyroglobulin (Tg), the most common marker to determine remission of differentiated thyroid carcinoma (DTC), can take 18 months or longer to be undetectable. We hypothesized that Tg stimulated after surgery and immediately before radioiodine treatment (baseline-stimulated Tg) could be a good predictor of remission at 18-24 months. The aim of this study was to evaluate the role of baseline-stimulated Tg as early prognostic marker of DTC. Patients and methods Retrospective study of 133 patients with DTC from 1998 to 2010 (age at diagnosis 47·4 ± 16·8, follow-up 5·09 ± 3·2 years). Initial subset analysis was performed after excluding patients with positive TgAb, who were later included in the second. Baseline-stimulated Tg was divided into tertiles. Multivariate logistic regression analysis included baseline Tg and other known prognostic markers and receiver operating characteristic (ROC) curve to identify the best cut-off level of baseline Tg were performed. Results Baseline-stimulated Tg in the highest tertile was the only predictive variable of persistence of disease at 18-24 months in the initial analysis (OR 45·3, P < 0·01). In the second analysis, the predictive variables were baseline-stimulated Tg (OR 39·6, P < 0·001), presence of TgAb (OR 23·4, P < 0·005) and uptake outside of the thyroid bed post-treatment whole body scan (WBS; OR 5·3, P < 0·05) were predictive of persistence of disease. The ROC curve showed that baseline-stimulated Tg below 8·55 μg/l identified 95% of disease-free patients at 18-24 months after initial treatment. Conclusions Baseline-stimulated Tg is a good predictor of remission of disease at 18-24 months after initial treatment and could be a useful marker to stratify risk immediately after surgery. © 2013 John Wiley & Sons Ltd.