Executive dysfunction in children with attention deficit hyperactivity disorder (ADHD) is thought to be closely related to the prefrontal cortex (PFC). However, there is controversy over the activation of the PFC in children with ADHD. Differences could be related to the subtype. Meanwhile, no study to date has used functional near-infrared spectroscopy (fNIRS) to explore the differences between subtypes. Thus, this study aimed to explore the activation of the PFC in children with different subtypes of ADHD during executive function task. Participants in this study include typically developing (TD) children (n = 28), ADHD-predominantly inattentive (ADHD-PI) (n = 39) and ADHD-combined (ADHD-C) (n = 24). To examine the executive function of ADHD, the Go/No-go task is chosen to assess the response inhibition function. The activation of PFC in all participants during the Go/No-go task was recorded by fNIRS. Meanwhile, behavioral data were also recorded. Both TD and ADHD children activated the right PFC [middle frontal gyrus (MFG) and inferior frontal gyrus (IFG)] during response inhibition. However, the range and degree of activation differed among these groups. Compared with TD children, those with ADHD-PI had a smaller extent of activation in the right PFC, and those with ADHD-C only had a tendency to enhance activation. In addition, children with ADHD-PI and ADHD-C had impaired activation of the temporal gyrus. Besides, compared with ADHD-C and TD, those with ADHD-PI also had impaired activation of the right precentral gyrus (PG), and the supplementary motor area (SMA). Compared with ADHD-PI, ADHD-C showed decreased activation of the right MFG. The activation of Ch34 (BA44, rPFC) in children with ADHD-PI and ADHD-C was negatively correlated with their clinical symptoms. The activation of the PFC in children with different subtypes of ADHD has both commonalities and differences. The degree of activation of the right PFC Ch34 in children with ADHD is negatively correlated with clinical symptoms. fNIRS could be served as a candidate hemodynamic biomarker for the diagnosis of ADHD.