OBJECTIVE To determine which clinicopathological factors influence the prognosis of cervical adenocarcinoma. METHODS Three hundred and two cases of primary adenocarcinoma of the uterine cervix, treated between 1977 and 1994, were studied retrospectively. Clinical data and pathological findings with respect to primary therapy were reviewed and evaluated. RESULTS The 5-year survival rates for stages I, II, and III/IV were 75.9, 62.9, and 25.1%, respectively. International Federation of Gynecology and Obstetrics stage (P < 0. 0001), cell type (P = 0.0176), tumor grade (P = 0.023), lymph node status (P = 0.018), and bulky tumor (P = 0.007) were found to be independent factors using the stepwise Cox proportional hazards model. Old age (P = 0.0581), presence of hypertension (P = 0.46), diabetes mellitus (P = 0.18), obesity (P = 0.15), and oral contraceptive use (P = 0.42) were not found to adversely influence survival rates for cervical adenocarcinoma after adjusting for other covariates. Adenosquamous adenocarcinoma had a better prognosis than endocervical columnar cell adenocarcinoma in stages I and II (P = 0. 0235). Also, in cervical adenocarcinoma's early stages, multivariate modeling revealed that chances of survival were significantly better for patients treated by radical surgery than for patients treated by radiation therapy (P < 0.001). CONCLUSIONS Survival rates for cervical adenocarcinoma were significantly influenced by stage, histologic subtype, tumor grade, the presence of a positive lymph node, and tumor size. Although a randomized prospective study is needed, our data imply that radical surgery may be considered a better primary modality of treatment than radiation therapy for the early stages of cervical adenocarcinoma. Further, the presence of hypertension, diabetes mellitus, or obesity may not adversely influence survival rates.