Use of postmenopausal hormone replacement therapy (HRT) has been hypothesised to be associated with a reduced risk of non-Hodgkin's lymphoma (NHL), but the epidemiologic evidence is conflicting. To examine the risk of NHL in HRT users aged 40 and older, we conducted a cohort study in the County of North Jutland, Denmark (population 0.5 million) using data from population-based health registries for the period 1989-2002. We computed age-standardised NHL incidence rates and used Cox regression analysis to compute the relative risk (RR) and corresponding 95% confidence intervals (CI) of NHL among HRT users compared with non-users, adjusting for age and calendar period. The number of prescriptions redeemed (1, 2-4, 5-9, 10-19, or 20 or more prescriptions) was used as a proxy for duration of HRT. We identified 40 NHL cases among HRT users during 179 838 person-years of follow-up and 310 NHL cases among non-users during 1 247 302 person-years of follow-up. The age-standardised incidence rates of NHL were 25.7 per 100,000 among HRT users and 24.2 per 100,000 among non-users, yielding an adjusted RR of 0.99 (95% CI: 0.71-1.39). Our data did not support an association between HRT use and risk of NHL.