Background: Little is known on the participation of immunoglobulin isotypes and subclasses in the pathogenesis of the severe disease caused by the pandemic influenza virus (influenza A(H1N1)pdm09). Objectives: (1) To evaluate the association between plasma levels of IgG1, IgG2, IgG3, IgG4, IgA, IgM, IgE and outcome in patients with severe pandemic influenza. (2) To evaluate the association between immunoglobulin and cytokine levels in these patients. Study design: 40 critically ill patients with community acquired pneumonia and influenza A(H1N1)pdm09 infection were recruited from November 2010 to February 2011. Plasma samples were collected during the first 24. h following admission to the ICU. Immunoglobulins and 17 major cytokines were profiled in plasma. Results: 15 patients died (37.5%). When the association between clinical variables and prognosis was assessed, prior immunosuppression, APACHE II score, levels of IgG2 and levels of IgM were associated with outcome in a univariate Cox regression analysis. Kaplan Meier analysis showed that patients with levels of IgG2 and IgM. <. 59 and. <. 58. mg/dl respectively died earlier. Multivariate Cox regression analysis showed that APACHE II score and levels of IgM were the best predictors of outcome, being levels of IgM a protective factor against mortality. IgM was the immunoglobulin showing the largest number of negative correlations with cytokine levels. Conclusions: Our results support a central role of IgM in preventing uncontrolled inflammatory response and mortality in severe pandemic influenza. Early assessment of IgM could contribute to guide clinical decisions in these patients. © 2013 Elsevier B.V.