Some studies suggest that transfusion may be associated with an increased risk of recurrence of and death due to malignant human neoplasms. We examined retrospective data from patients with cervical cancer to see if any association between transfusion of blood at the time of initial treatment and the time interval to recurrence and death could be detected in this cancer. In 130 patients with cervical cancer, seen over a ten-year period at our institution, there was a trend toward earlier recurrence in transfused patients, but this trend did not achieve statistical significance. Death due to cervical cancer recurrence occurred after a median of Death due to cervical cancer recurrence occurred after a median of 12 months in the transfused patients and a median of 68 months in the non-transfused individuals, which was statistically significant. Transfused patients had, on average, more favorable prognostic factors than those not transfused, such as less advanced clinical stage of disease. Analysis using a proportional hazards risk model failed to demonstrate a significant association between transfusion and time to recurrence when other prognostic factors were considered, but a significant association between transfusion and time to cancer-related death (p less than 0.05) was found. While these results cannot be viewed as conclusive due to the small number and heterogeneity of the patients analyzed, our data support the possibility of an association between transfusion and cervical cancer survival. Further studies are warranted to confirm or refute this relationship.