Age at diagnosis has been proven to be an important determinant of the choice of initial treatment for several sites of cancer. Elderly patients are more likely to receive no treatment or less intensive treatment modalities. This study analysed the influence of age on treatment choice and survival in patients diagnosed with cervical cancer. This population-based study used data on 1176 new cases of invasive cervical cancer diagnosed in the period of 1986-1996 from three regional cancer registries in the Netherlands. All available information on treatment and survival (on 1 January 1998) was recorded. Relative survival rates were calculated according to the Hakulinen method. Relative risks (RR) for excess mortality due to the diagnosis of cervical cancer were calculated with a regression model for relative survival rates. Only 5% of the patients aged 70 years and older (n=224) were diagnosed with stage IA disease, compared with 11 and 30% of the patients aged 50-69 years and 49 years and younger, respectively. Almost 50% of the 70+ patients with stage IB-IIA were treated with radiotherapy as a single treatment modality, whereas 64% of the patients aged < or =49 years were treated with surgery alone. In all age groups, treatment for advanced stage disease (stage > or =IIB) was radiotherapy alone. No treatment was given to 10% of the patients aged 70 years and older, 5% of those aged 50-69 years and 1% of those aged 49 years and younger. Five-year relative survival was 69% (95% Confidence Interval (CI): 66-72%) and differed significantly (P=0.001) with age (70+ years: 49%; 50-69 years 58%; < or =49 years: 81%). Multivariate analyses on a subset of patients showed that age was not an independent prognostic factor, whereas stage and treatment modality were very important prognostic factors. Although elderly cancer patients were sometimes treated differently from younger patients, this was in accordance with the guidelines. Relative survival rates differed significantly by age. The multivariate analyses on the subset of patients also revealed that excess mortality increased with age. However, when adjustment was made for stage and treatment, this difference disappeared. The influence of treatment on survival is likely to be due to the selection of patients based on other characteristics, such as tumour volume, comorbidity and performance status.