The aim of this study was to assess the feasibility of using objective data obtained at diagnosis of Hodgkin's disease to predict those patients who were likely to die of progressive disease within 4 years of diagnosis. 92 consecutive patients from one centre (Newcastle upon Tyne) were used to construct a numerical index based on disease stage (Ann Arbor), age, haemoglobin and absolute lymphocyte count. Weight was assigned according to a predictive value in univariate and multivariate analyses based on survival. The index produced was then validated on a separate patient set (455) from other centres within the Scotland and Newcastle Lymphoma Group (SNLG) on whom the same prospective information was available. The index produced provided a useful separation of those patients destined to die of disease. Of 101 patients with index higher than 0.5, 62 (61.4%) were dead at 4 years, whereas with index lower than 0.5, 61 (18%) of 336 patients were dead at 4 years. The index includes Ann Arbor stage but possesses additional practical prognostic value which allows identification of patients with early stage destined to die of disease. Of 149 patients with stage IA and IIA disease 15 patients had index higher than 0.5, and 10 (60%) have died, whereas the remaining patients had survival of 90% and 85% respectively. This numerical index is applicable to all patients at diagnosis and in the SNLG population gives better predictive survival at 4 years than stage alone, and provides a basis for selecting patients for more aggressive therapy.