The National Asthma Education and Prevention Program (NAEPP) Expert Panel II recommended a stepped care pharmacotherapy approach to asthma treatment based on an objective assessment of asthma severity using daytime symptoms, nocturnal symptoms, and physiologic lung function. The worst grade of the individual variables determines overall asthma severity. With this approach, patterns of asthma severity categorization might vary among individual variables; one variable might have a predominant effect on overall categorization. During the run-in, pretreatment phase of five controlled clinical trials, data from 744 inhaled steroid nonusers and 685 inhaled steroid users on asthma control were collected and asthma severity categorized. In inhaled steroid nonusers nocturnal symptoms classified the majority of patients as severe, persistent, but wheeze classified 27.3% of patients as mild, intermittent and 25.7% as mild, persistent. If the worst grade from the four asthma symptoms was used for severity grading, most patients were categorized as severe, persistent. beta-Agonist use and FEV(1) classified most as moderate, persistent. There was poor correlation between variables in severity categorization. Severity grading for European patients was similar to that for U.S. patients. Applying the Expert Panel II recommended method for asthma severity categorization to a large data set illustrates that a single variable, nocturnal symptoms, determined to a large extent overall categorization. Development of a validated method for asthma severity categorization is essential for using a stepped care approach to asthma pharmacotherapy.