Following the recent debates on the discrepancy between the predominant weight of bipolar disorder (BPD) in the clinical reality and its relatively low prevalence figures emerging from epidemiological surveys, the present paper contends the ability of current operational diagnostic system to properly detect the clinical entity of bipolar disorder.As an episode of mania/hypomania is the necessary requirement for a diagnosis of bipolar disorder to be made, in this editorial we maintain that: a) the most severe forms of mania, characterized by cloudy consciousness, mood incongruent delusions, and physical symptoms are likely to escape DSM IV criteria, that are shaped around hypomania or mild mania; b) the impossibility to diagnose mania when this occurs during antidepressant treatments impedes diagnosing those cases whose natural illness pattern is Depression followed by Mania (known as DMI pattern); c) given that approximately 50% of cases have their onset of BPD with affective episodes other than mania/hypomania any prevalence figure necessarily underestimates BPD; d) the sub-threshold forms of BPD, well described in the concept of Bipolar Spectrum, are beyond the possibility to be recognized using operational diagnoses in spite of their utmost clinical relevance.