Background: Exercise training improves cardiometabolic outcomes in ‘mean terms’, but little information is available in children about the impact of the frequency/week and the wide inter-individual variability to exercise training reported in adults. Objectives: We compared the effects of resistance training (RT) and high-intensity interval training (HIT), and ‘high’ and ‘low’ frequency of training/week, for their effectiveness in decreasing insulin resistance (IR) levels in schoolchildren. A second aim was to decscribe and compare the prevalence of non-responders (NRs) between the different frequencies of training protocol. Methods: Fifty-three schoolchildren with IR were randomly assigned into four groups: RT at high frequency (three times/week), HIT at high frequency, RT at a low frequency (two times/week) and HIT at low frequency. The intervention lasted 6 weeks. Blood samples and body composition, blood pressure and performance measurements were taken before and after the intervention. Results: The prevalence of NRs was similar between the RTHF and HITHF (25.0% vs. 25.0%, P > 0.05) and RTLF and HITLF groups (20.0% vs. 46.6%, P = 0.174) for decreasing homeostasis model assessment of IR. However, significant differences in the prevalence of NRs were detected between RTHF and HITHF groups in fasting glucose (FGL) (18.7% vs. 58.3%, P < 0.031). Conclusions: Both RT and HIT improves the glucose control parameters in schoolchildren over 6 weeks, but only HIT is independent of a high or low frequency of training/week. The prevalence of NRs is similar for decreasing homeostasis model assessment of IR comparing each exercise mode in high vs. low frequency/week. However, both high- and low-frequency RT and HIT results in differences in the prevalence of NRs for FGL and other cardiometabolic and performance outcomes.