In general, elderly patients with aggressive non-Hodgkin's lymphoma (NHL) have a less favourable prognosis than younger patients. Established predictors of prognosis in NHL are less discriminatory in the elderly, which is why there is a need for additional markers giving guidance on treatment decisions and prediction of outcome. The expected length of life of an individual in the general population is intimately associated with that of his/her parents. The aim of this study was to test the hypothesis that parental longevity is associated with improved outcome also among elderly patients with aggressive NHL and thus serves as an easily accessible non-disease associated prognostic factor. A total of 220 patients ( > 60 years) with aggressive NHL with a median age of 71 years (range 60-86) were included. Patients were randomized to receive CHOP or CNOP (doxorubicin replaced with mitoxantrone) chemotherapy with or without the addition of granulocyte colony-stimulating factor. The median follow-up time was 56 (19-89) months. Parental data regarding age at death were available through parish offices for 425 (97%) parents. Relative risk (RR) of death (disease-specific and all-cause) associated with parental lifespan was assessed using Cox proportional hazards regression analyses, with adjustment for sex, age, prognostic index, symptoms, and calendar period of diagnosis. Maternal lifespan below (versus above) median was associated with a borderline significant reduced disease-specific (adjusted RR of death from NHL = 1.5; 95% confidence interval 1.0-2.1) and overall survival. The effect of maternal lifespan was somewhat more pronounced in patients receiving CHOP than CNOP treatment. Paternal lifespan below the median was associated with a borderline significant increased disease-specific (adjusted RR of death from NHL = 0.8 [0.5-1.0]) and overall survival. Combined, maternal, and paternal lifespan had little impact on survival. These effects were true also when CHOP and CNOP treated patients were analysed separately. Maternal and paternal lifespan may predict survival in NHL, but with opposing effects. At present parental age appears not to be a clinically useful predictor of prognosis in the elderly with aggressive NHL.