8550 Background: Hodgkin lymphoma (HL) is a heterogeneous disease, but differences between nodular lymphocyte predominant (NLPHL) and classical (CHL) subtypes were previously studied in small cohorts preventing adjustment for confounders. We studied those differences based on the Surveillance, Epidemiology and End Results (SEER) program data. Methods: We analyzed SEER HL cases aged 16 years and over, diagnosed between 1995 and 2009. We studied the following endpoints: crude probability of HL-related death (HLRD), relative survival (using individual data in multivariate flexible parametric models) and risk of secondary malignancies (using competing risk regression). We studied the impact of radiotherapy (RT) in early-stage disease using a propensity score, adjusting for treatment selection and immortal time bias. Results: We identified 25,903 patients, with disparate age, race and stage distributions between subtypes. In a multivariate model, NLPHL demonstrated significantly better crude and net survival outcomes (Table), but in contrast to all CHL subtypes, it showed a steady increase in mortality rate after 2 years. The risk of secondary non-Hodgkin lymphoma was significantly higher in both NLPHL (hazard ratio, HR, 2.28, P=0.002) and lymphocyte-rich (LR) CHL (HR 2.08, P=0.01) than in nodular sclerosis (NS). The risk of other secondary malignancies did not differ between subtypes. After balancing confounding factors in treatment arms, RT in stage I/II was associated with improved survival in NS (HR, 0.78, P=0.001) and LR-CHL (HR 0.36, P<0.001), but not in NLPHL (HR 0.98, P=0.94) or in the remaining CHL subtypes. Conclusions: Studies of NLPHL and CHL subtypes should account for their disparate biology and clinical course. Prospective evaluation of NLPHL treatment strategies without RT is justified. [Table: see text]