Between 1962 and 1985, 371 patients had initial treatment for bulky endocervical carcinomas of the uterine cervix at The University of Texas M.D. Anderson Cancer Center. All patients had concentric expansion of the cervix by tumors that measured at least 6 cm in greatest transverse diameter. Of the 361 patients treated with curative intent, 211 (57%) had FIGO Stage I disease that was believed to be confined to the uterus, 59 (16%) had FIGO Stage IIA disease, and 101 patients (27%) had FIGO Stage IIB disease. Median follow-up time of surviving patients was 130 months. Actuarial survival rates of 361 patients treated with curative intent were 54% and 48% at 5 and 10 years, respectively. The actuarial pelvic disease control rate was 76% at 10 years. Patients whose tumors were less than 8 cm in maximum diameter (279 patients) had a better survival rate than those with tumors greater than or equal to 8 cm in diameter (92 patients) (p less than 0.01). Of the 282 patients who underwent lymphangiography, survival rate was significantly better for those with negative studies than it was for the 113 patients (40%) with positive or suspicious studies (p less than 0.01). There was no correlation between FIGO stage and survival rate (p = 0.64) or pelvic control rate (p = 0.59). Of patients treated with curative intent, treatment was by radiation alone (RT) in 244 (68%) or by radiation followed by hysterectomy (RT+S) in 117 (32%). Although there has been an overall shift in policy away from the use of adjuvant hysterectomy during the past decade, many patient selection factors also influenced the choice of treatment during the study years, resulting in a significantly higher proportion of patients with adverse prognostic features in the RT group. Patients chosen for treatment with RT alone had a greater likelihood of having tumors greater than or equal to 8 cm (p = 0.03), FIGO stage IIB (p less than 0.01), positive lymphangiogram (p = 0.02), and persistent palpable parametrial disease after external radiotherapy (p less than 0.01). Patients treated with RT alone also had a lower overall survival rate at 10 years than patients treated with RT+S (45% vs 64%, p less than 0.01). Although multivariate analysis suggested that treatment had an independent influence upon survival rate, it was difficult to draw firm conclusions about the value of adjuvant surgery because of the numerous biases in patient selection, some of which may have been difficult to quantify.(ABSTRACT TRUNCATED AT 400 WORDS)