BACKGROUND An increased risk of non-Hodgkin lymphoma (NHL) has been reported among long-term users of tricyclic antidepressants. METHODS The incidence of NHL among 43,932 users of any antidepressant medication (2 or more prescriptions) in the county of North Jutland, Denmark, during 1989-2003 was compared with NHL incidence in all those not using antidepressants (0 or 1 prescription). We used Poisson regression analysis adjusted for age, sex, period, use of immunosuppressive drugs, corticosteroids, or nonsteroidal anti-inflammatory drugs. RESULTS Use of tricyclic antidepressants was associated with an overall increased incidence of NHL (adjusted incidence rate ratio [IRR] = 1.53; 95% confidence interval [CI] = 1.06-2.21) compared with nonuse. The IRR in users with > or =10 prescriptions and 5 or more years of follow-up was 2.50 (CI = 1.43-4.34). Users of other types of antidepressants were not at increased risk of NHL. CONCLUSIONS Our results indicate an increased risk of non-Hodgkin lymphoma specifically among long-term users of tricyclic antidepressant medications.