Intermittent preventive treatment (IPT) with dihydroartemisinin-piperaquine (DP) is highly protective against malaria in children, but is not standard in malaria-endemic countries. Optimal DP dosing regimens will maximize efficacy and reduce toxicity and resistance selection. We analyze piperaquine (PPQ) concentrations (n = 4573), malaria incidence data (n = 326), and P. falciparum drug resistance markers from a trial of children randomized to IPT with DP every 12 weeks (n = 184) or every 4 weeks (n = 96) from 2 to 24 months of age (NCT02163447). We use nonlinear mixed effects modeling to establish malaria protective PPQ levels and risk factors for suboptimal protection. Compared to DP every 12 weeks, DP every 4 weeks is associated with 95% protective efficacy (95% CI: 84-99%). A PPQ level of 15.4 ng/mL reduces the malaria hazard by 95%. Malnutrition reduces PPQ exposure. In simulations, we show that DP every 4 weeks is optimal across a range of transmission intensities, and age-based dosing improves malaria protection in young or malnourished children.