Aerosolized medications can potentially be delivered to the lungs of infants through a nasal cannula interface. However, nose-to-lung delivery technologies currently allow for ~1% of the loaded dose to reach an infant’s lungs. Conventional dry powder inhalers (DPI) are superior to other types of inhalers in many ways. However, passive DPIs that operate based on user inhalation and require large volumes of airflow are not applicable to infants. To overcome this challenge, positive pressure DPIs have been developed that enable aerosol delivery to infants. Unless an adequate nasal interface is used with these devices, a significant amount of drug will still be lost. Computational fluid dynamics (CFD) provide a method to assess the performance of a nasal cannula interface and optimize its performance. In this study, a CFD model was first experimentally validated using the low-Reynolds number k-ω turbulence model, then used to assess and optimize several conical diffuser cannula designs for infants. The performance of a cannula depends primarily on two requirements: the amount deposited particles and the cannula’s volume. It was found that 90 and 100 mm long simple diffusers achieved the necessary deposition and volume requirements when operated at 3 and 5 liters per minute, respectively. Additionally, including clean sheath co-flow air with the 70 mm long diffuser achieved the targeted performance requirements. Inclusion of recent advancements in the field with the recommended cannula designs is likely to improve pharmaceutical aerosol delivery to infants using the nose-to-lung approach.