We investigated prognostic factors in 42 anaplastic thyroid carcinoma (ATC) patients from a single institution over a 30-year period and explored the use of risk grouping to guide therapeutic decisions. Univariable and multivariable differences in overall survival (OS) were evaluated using the Kaplan-Meier method and the log-rank test as well as Cox proportional hazards model. Risk grouping in making therapeutic decisions for ATC patients was explored. The 1- and 3-year OS rates were 28.6 % and 18.5 %, respectively. Univariate analysis indicated that 4 pre-therapeutic factors of patients were related to poorer prognoses: age ≥ 55 years, white blood cell count ≥ 10.0 × 10(9)/L, blood platelet count ≥ 300.0 × 10(9)/L and advanced clinical tumor-node-metastasis stage. These factors were used to calculate the risk indices. Patients with total risk index scores of no more than 1 were considered to be in the low-risk group, and patients with scores ≥ 2 were considered to be in the high-risk group. The patients in the low-risk group had significantly better 1- and 3-year OS rates (90.9 % and 63.6 %, respectively) than those in the high-risk group (6.5 % and 3.2 %, respectively). Risk group and therapeutic regimen were the 2 factors that independently influenced survival according to multivariable analysis. Surgery that was combined with postoperative radiotherapy significantly benefited the patients in the low-risk group rather than the patients in the high-risk group. Risk grouping was a helpful tool of evaluating the prognoses and guiding the treatment of ATC patients.