The importance of the subclassification of nodular sclerosing Hodgkin's disease (NSHD) according to the British National Lymphoma Investigation Group (BNLI) criteria, which had an independent prognostic value for 90 unselected patients diagnosed in our region in the period 1972-83, is still equivocal. Because survival of patients with Hodgkin's disease improved in our region during the period 1972-92 and because treatment may modify prognostic factors, we re-evaluated the prognostic value of NSHD subclassification up to 1993. A registry-based study was performed with data on all 345 Hodgkin patients diagnosed in the period 1972-92. Available histology was reviewed. The prognostic value of nodular sclerosis (NS) grading was evaluated for two periods, 1972-80 and 1981-92, designated as the seventies and the eighties, respectively. NSHD was diagnosed in 57% (n = 195) of all registered cases of Hodgkin's disease, 17 of which could not be evaluated. NS stage I (NSI; 73%) and NS stage II (NSII; 27%) patients exhibited the same distribution of stage during both periods; NSII patients were older than NSI patients in the seventies. NSII patients presented with an elevated ESR and B symptoms more frequently during the eighties and subsequently received combined modality therapy more often. The crude 5-year survival rate for grade I v grade II NSHD was 85% v 38% (P < 0.05) for the seventies and 84% v 83% for the eighties. Subclassification of NSHD was not an independent prognostic factor after adjustment for age, stage, gender, B symptoms and ESR, though it remained of independent prognostic value when ESR was left out of the Cox model. The most important factors adversely influencing survival were advanced age, advanced stage and male gender. The independent prognostic value of the subclassification of NSHD has disappeared, as reflected by the clearly improved survival of NSII patients, probably due to more intensive treatment in the eighties, whereas survival of NSI patients did not changes.