In the 1970s the introduction of regular monitoring of peak expiratory flow rate (PEFR) in asthma patients led to the recognition of patterns of variation that had clinical importance. Most noticeable was severe “morning dipping” or diurnal variation which was shown to be a prominent feature in small numbers of patients shortly before death or successful cardiopulmonary resuscitation from apparently suddenly worsening asthma.1 ,2 This phenomenon has subsequently been attributed to the effects of normal circadian rhythms in airway calibre on increasing bronchial hyperreactivity3and the underlying mechanisms are at least partly understood. Turner-Warwick also described another pattern, “brittle asthma”, which showed similar large changes in the magnitude of PEFR, responsiveness to frequent use of inhaled bronchodilators, but a chaotic pattern of variation.4 Some authors suggested that asthma patients could die within very short periods of apparently good asthma control,5 although others debated whether these deaths were only perceived to be sudden through poor supervision.6 Twenty years on these phenomena remain an enigma. Two papers in this issue of Thorax seek to define these conditions more precisely and determine their true nature.

Ayres et al 7 have reviewed current thinking on brittle …