The emergence of novel, effective but expensive biological therapies is significantly improving outcomes for many patients with cancer, but also substantially increasing treatment costs. Monoclonal antibodies with long half-lives have often been developed using loading doses to hasten the achievement of a therapeutic dose. Although the benefits of loading doses have not been proven, we caution against abandoning this practice on uncertain theoretical grounds. Rather, the issue of loading doses should be seen in the broader context of how best to define the optimal dose, schedule and duration of treatment through novel clinical trial designs.