SummaryThe survival of patients diagnosed with Hodgkin lymphoma (HL) has improved from 70% to 90% in clinical trials. However, population-based data has shown lower survival. In this study, clinical trial data were linked with cancer registry to identify trial and non-trial participants and differences in overall survival and associated factors were assessed. In 1986-2004, 27% of HL patients aged 15-70 years participated in clinical trials. Compared to non-trial participants, trial participants were younger (median age, 31 vs. 34 years), had staging registered more accurately and had an 8% higher 20-year survival rate (73% vs. 65%). After adjusting for baseline differences, no differences in survival (hazard ratio = 0Á96, 95% confidence interval 0Á82-1Á12), or in subgroup analysis according to stage, remained. Over time, increased administration of chemotherapy in combination with radiotherapy, together with the decreased use of radiotherapy alone was observed among the trial population. This trend was later followed in nontrial participants, coinciding with a similar 'take-up' in survival. The observed superior survival among patients with HL treated in clinical trials can be largely explained by the differences in baseline characteristics, particularly younger age. High trial participation rate and centralized expertise facilitates the implementation of trial findings to real-world practice.Keywords: overall survival, hodgkin lymphoma, clinical trials, populationbased cancer registry, deterministic data linkage. Between 1964 and, the Lymphoma Group of the European Organisation for Research and Treatment of Cancer (EORTC) has conducted a series of phase III randomized clinical trials (RCTs) aiming to improve the efficacy of treatment in Hodgkin Lymphoma (HL), while minimizing longterm treatment-related toxicities. These subsequent trials have successfully improved the 10-year relative survival (RS) of HL patients, diagnosed between 1964 and 2004, with early stage disease from 72% to 93% and with advanced staged disease from 63% to 78% (Favier et al, 2009). However, compared to the high survival rate observed in the most recent clinical trials, that of the general patient population seems to be lower. In the Netherlands, for instance, for patients diagnosed with HL between 1989 and 2000, the 5-year RS was 78% in the general patient population (including trial participants), which is much lower than the 5-year RS seen in patients entered in clinical trials (87%). With longer follow-up, the survival gap between these two patient research paper ª