The aims of this study are to identify prognostic factors of differentiated thyroid carcinoma and to validate the application of prognostic classifications obtained by others studies to another population. The survival of 208 patients with differentiated thyroid carcinoma (129 papillary and 79 follicular carcinoma) was calculated by the Kaplan-Meier method. The mean follow-up was 7.5 years (1-17.7). Cox-proportional hazard model was used for variables influencing on survival (Mantel-Cox method). In addition, the EORTC, AGES, AJC, AMES and DeGroot classifications were tested. The independent prognostic factors identified were patient age > 60 years, tumor size > 6 cm and the presence of distant metastases. The absence of poor prognostic factors defined the low risk group (153 patients, survival 97% at 17.7 years of follow-up). The application of the other prognostic classifications differentiated several risk groups not in accordance with those obtained in the initial population. In this series of patients with thyroid carcinoma the main prognostic factors were age, tumor size and the presence of distant metastases. However, there are pitfalls in applying the prognostic classifications of published studies to another population.