This study aimed to compare the efficacy of methylphenidate and atomoxetine on improving executive functions among children with attention-deficit/hyperactivity disorder (ADHD). This was an open-label, head-to-head, 3-month, randomized clinical trial with two-arm parallel-treatment groups: osmotic-release oral system methylphenidate (OROS-MPH;  = 79) and atomoxetine once daily ( = 78). Three major domains of executive functions were assessed, including response selection/inhibition, flexibility, and planning/working memory. The neuropsychological measures included the Conners’ continuous performance test and the Cambridge Neuropsychological Test Automated Battery. We found that both treatment groups showed improvement in executive functions (-value <0.05 for the major indices of each domain). In addition, OROS-MPH was associated with a greater magnitude of improvement in the response selection/inhibition; the slope for detectability improvement in the Conners' continuous performance test was 0.06 for atomoxetine and 0.15 for OROS-MPH (-value <0.01); the slope in rapid visual information processing was 2.22 for atomoxetine and 3.45 for OROS-MPH (-value <0.05). Both OROS-MPH and atomoxetine improved various domains of executive functions in children with ADHD. There is greater improvement in response selection/inhibition among patients treated with OROS-MPH than those with atomoxetine. This trial was registered with ClinicalTrials.gov (no. NCT00916786).