A retrospective, case-control analysis of 30 women with cervical cancer associated with pregnancy, surgically managed at the University of Iowa between 1960 and 1994, was performed. Controls were matched with cases based on age, histology, stage, treatment, and year of treatment. Patients were divided into two groups: Group I, radical hysterectomy (26 patients) and group II, simple hysterectomy (4 patients). Eleven patients underwent surgical treatment in the third trimester with a mean planned delay in therapy of 16 weeks. None of the patients with a planned delay in therapy developed recurrent disease. No neonatal morbidity was encountered in these patients. Among group I patients, there was longer anesthesia time (P < 0.03), but there were no differences in the mean operative time. There was more blood loss at the time of surgery among pregnant patients (1493 cc vs 1065 cc for group I, P = 0.005; 812 cc vs 362 cc for group II, P = 0.03); however, there was no difference in the frequency of blood transfusion. The percentage of patients receiving a transfusion decreased significantly after 1991 (33% versus 90%, P = 0.01 for pregnant patients and 33% versus 85%, P = 0.03 for nonpregnant patients). There were no differences in the time required for postoperative bladder drainage, mean hospital stay, febrile morbidity, incidence of wound infection, wound separation, pelvic abscess, thromboembolic disease, or urinary tract infection. One case patient and 3 control patients died of disease, but this difference was not statistically significant. Based upon our data, in selected cases of early-stage cervical cancer, surgical management of cervical cancer is safe during pregnancy. For early Stage I squamous cancers, planned delay in therapy is safe.