Psychiatric disorders, including attention-deficit/hyperactivity disorder (ADHD), are diagnosed according to standard diagnostic criteria—such as the Diagnostic and Statistical Manual of Mental Disorders (DSM-V)—primarily based on perceivable traits. Patients displaying similar perceivable traits are categorized with the same diagnosis. Consequently, a wide variety of patients with different symptoms and psychosomatic and physiological etiologies of the disease share the same diagnostic classification (Buyck et al., 2014; Loo et al., 2013; Martel et al., 2011). This heterogeneity of patients results in the limited therapeutic efficacy of medications in psychiatric disorders (Blouin et al., 2010; Ter-Stepanian et al., 2010). ADHD is also known to have a response rate of approximately 70% for pharmaceutical preparations, which is higher than those of other psychiatric disorders, but it is not a small population of ADHD patients who are nonresponders to any treatment (Duong et al., 2012; Spencer et al., 2005; Biederman and Spencer, 2008). While the determining factors for good or poor responses to pharmaceutical preparations has not yet been entirely elucidated, we hypothesized that physiological brain activities play an important role in this characterization. electroencephalography (EEG) provides a noninvasive approach to evaluating human brain activity in an objective manner. As such, EEG is potentially suitable for characterizing the heterogeneous properties of patients suffering from psychiatric disorders such as ADHD. To date, a considerable number of studies have reported EEG abnormalities in ADHD patients in band power (Barry et al., 2003, 2010; Chabot and Serfontein, 1996; Clarke et al., 2001c, 2008; Dupuy et al., 2014; Koehler et al., 2009; Lazzaro et al., 1998; Olbrich et al., 2016), and band power ratio (Angelidis et al., 2016; Monastra et al., 2001). Since cognition and behavior are thought to be based not only on the activity of a single area but also on the integrated activity between different areas, several coherence abnormalities of ADHD patients have been also reported (Barry et al., 2002; Clarke et al., 2007b). Consequently, the FDA has approved the theta/beta (θ/β) ratio as an EEG marker recommended to support the diagnosis of ADHD. However, most preexisting EEG markers do not necessarily fluctuate as a reflection of the treatment effects (Loo and Barkely, 2005). Some researchers have tried to explore predictive biomarkers for preexisting medications (Arns et al., 2008, 2018; Barry et al., 2009; Clarke et al., 2002a, 2002b, 2005, 2007b; Ogrim et al., 2014; Olbrich et al., 2013), but there have not yet been any certified good predictive markers for each medication. The absence of such predictive biomarkers not only imposes difficult choices regarding prescriptions on doctors but also makes the development of novel drugs for patients who are non-responders to preexisting agents difficult. We hypothesize that the segmentation of ADHD patients by treatment efficacy and comprehensive exploration of EEG phenotypes (band power and coherence analysis for broad band range) can elucidate a heterogeneity of ADHD patients and leads to identify novel predictive markers for each medication. In this study, we confirmed the transition of widely recognized ADHD specific EEG abnormalities before and after treatment. Moreover, we divided ADHD patients into a treatment-responder and a non-responder based on the amount of change in ADHD rating scales (ADHD-RS), and attempted to identify EEG indices that could be used to predict the treatment efficacy.