Results of epidemiologic studies on postmenopausal hormone (PMH) use and non-Hodgkin lymphoma (NHL) are inconsistent. To help clarify this issue, PMH and NHL incidence was examined in the Cancer Prevention Study-II Nutrition Cohort. Between 1992 and 2007, 616 cases of NHL were identified among 67 980 postmenopausal women who were cancer-free at baseline. PMH use was updated during follow-up. Using extended Cox regression, we observed a statistically significant 29% higher risk of NHL for ever unopposed estrogen use compared to never use, which was restricted to follicular lymphoma (current estrogen compared to never use, hazard ratio [HR] = 2.25, 95% confidence interval [CI]: 1.17-4.33) and diffuse large B-cell lymphoma (DLBCL, HR = 1.95, 95% CI: 1.13-3.35). There was no association between current estrogen plus progestin (E + P) use and NHL incidence overall, but a suggested positive association between current E + P use and DLBCL, as well as former E + P use and follicular lymphoma. These results suggest that postmenopausal hormones might play a role in NHL etiology, particularly for follicular lymphoma and DLBCL.