The use of inhaled corticosteroids (ICS) is one of the most controversial issues in COPD treatment. There is evidence that ICS with or without long-acting beta-2-adrenergics (LABA) reduce exacerbation rates and improve the health status of severe COPD patients. The effects on FEV1 are limited and their effect on survival is unknown. Recently, a worldwide double-blind placebo-controlled study 'Towards a revolution in COPD health' (TORCH) was published in which salmeterol 50 microg and fluticasone 500 microg twice daily was compared with placebo, salmeterol alone or fluticasone alone. Of the 6112 included patients 875 died within 3 years after the start of treatment. Death rate, defined as the primary outcome parameter, was not significantly reduced (p = 0.052) while exacerbation frequency and health status improved in the combination therapy group. The TORCH trial therefore does not change current guidelines regarding trial treatment with ICS in severe COPD patients.