Treatment of sleep apnea patients may be justified even when symptoms are absent. On the one hand, hypersomnia or daytime sleepiness are subjective symptoms and only reflect one aspect of the clinical syndrome of sleep apnea. On the other hand, an apnea index in excess of 5, e.g. the "laboratory diagnosis" of sleep apnea, has been reported as an independent risk factor for myocardial infarction. The main problem is scientific definition of cut-off points for treatment. There is no controversy about whether patients with symptoms should be treated or not. A CPAP trial seems to be justified in asymptomatic patients with an apnea/hypopnea index in excess of 20-30 and in patients with cardiovascular risk factors when the apnea/hypopnea index is in excess of 5. These laboratory cut-off values are not absolute values but represent flexible guidelines for initiating a CPAP trial in asymptomatic patients with apnea. The reasons for uncertainty whether or not to treat patients with asymptomatic sleep apnea are discussed.