Cancer drug researchers need Phase II designs that balance cost and speed with scientific rigor and predictive value. Concurrent controls have come into greater usage recently in Phase II oncology, but there is still reluctance to use concurrent controls since randomized controlled trials (RCT) often require greater time and cost to enroll an adequate number of patients compared to single-arm designs. In this talk, we present examples of Phase II designs that attempt to be more efficient in their use of control information than standard RCTs. The first case study illustrates the Bayesian Augmented Control design, in which relevant historical data is used to augment a concurrent control arm using a Bayesian hierarchical model. The second case study demonstrates the Shared Control design, which allows multiple experimental treatments to be studied using a single concurrent control arm.