Investigators involved in the development of cancer therapeutics are testing new trial designs and endpoints in order to accommodate the perceived challenges in defining appropriate doses and schedules for further testing. Many new agents with specific molecular targets have entered clinical development or are being considered for development. While some of the agents have both toxicity and antitumour efficacy apparent at clinically achievable doses, thus the use of traditional algorithms is appropriate, others have significant clinical activity at doses considerably lower than the maximum tolerated dose. New initiatives in clinical trial design, both phase I and phase II may allow the development of appropriate plans for the development of these new molecularly targeted agents. Measures of target effect (tissue or imaging) are now commonly included in early trials of new targeted compounds, in an attempt to demonstrate proof of principle as well as guide dose selection. Phase II trial designs including novel correlative, imaging and clinical endpoints are being tested. Alternate endpoints such as progression or time to progression are being increasingly considered, and novel designs such as randomized discontinuation designs, multinomial designs and growth modulation indices are being prospectively tested. Progress in this area of early trial design are reviewed.