BACKGROUND Previous studies have provided limited support to the association between tobacco smoking and lymphomas with weak evidence of a dose-response relationship. METHODS We investigated the relationship between tobacco smoking and risk of non-Hodgkin lymphomas (NHL) and Hodgkin lymphomas (HL) through logistic regression spline models. Data were derived from an Italian hospital-based case-control study (1999-2014), which enrolled 571 NHLs, 188 HLs, and 1004 cancer-free controls. Smoking habits and other lifestyle factors were assessed through a validated questionnaire. Odds ratios (OR) and 95% confidence intervals (CI) were estimated by logistic regression, adjusting for potential confounders. RESULTS Compared to never smokers, people smoking ≥15 cigarettes/day showed increased risks of both NHL (OR = 1.42, 95% CI: 1.02, 1.97) and HL (OR = 2.47, 95% CI: 1.25, 4.87); the risk was particularly elevated for follicular NHL (OR = 2.43; 95% CI:1.31-4.51) and mixed cellularity HL (OR = 5.60, 95% CI: 1.31, 23.97). No excess risk emerged for former smokers or people smoking <15 cigarettes/day. Spline analyses showed a positive dose-response relationship with significant increases in NHL and HL risks starting from 15 and 21 cigarettes/day, respectively, with the most evident effects for follicular NHL and mixed cellularity HL. Smoking duration was significantly associated with the HL risk only (OR = 2.15, 95% CI: 1.16, 3.99). CONCLUSIONS These findings support a role of tobacco smoking in the etiology of both NHL and HL, providing evidence of a direct association of risk with smoking intensity.