Ventilation heterogeneity is an important marker of small airway dysfunction in asthma. The frequency dependence of respiratory system resistance (Rrs) from oscillometry is used as a measure of this heterogeneity. However, this has not been quantitatively assessed or compared with other outcomes from oscillometry, including respiratory system reactance (Xrs) and the associated elastance (Ers). Here, we used a multibranch model of the human lung, including an upper airway shunt, to match previously reported respiratory mechanics in mild to severe asthma. We imposed heterogeneity by narrowing a proportion of the peripheral airways to account for patient Ers at 5 Hz, and then narrowed central airways to account for the remaining Rrs at 18 Hz. The model required >75% of the small airways to be occluded to reproduce severe asthma. While the model produced frequency dependence in Rrs, it was upward-shifted below 5 Hz compared with in-vivo results, indicating that other factors, including more distributed airway narrowing or central airway wall compliance, are required. However, Ers quantitatively reflected the imposed heterogeneity better than the frequency dependence of Rrs, independent of the frequency range for the estimation, and thus was a more robust measure of small-airway function. Thus, Ers appears to have greater potential as a clinical measure of early small-airway disease in asthma.