Abstract The Methodology for the Development of Innovative Cancer Therapies (MDICT) task force considered aspects of the design and conduct of early (phase I and II) studies of combinations of molecular targeted agents during their 2012 meeting. The task force defined necessary non-clinical data, such as evidence of additive or synergistic effects in multiple molecularly credentialed and validated models, and appropriate pharmacodynamic marker development. A robust hypothesis was considered critical while non-clinical pharmacokinetic studies were also considered valuable. Clinical trials should include clear objectives that will prove or disprove the hypothesis. Predictive biomarkers/classifiers should be explored in phase I studies, rather than used to select patients. Trial design should be efficient and flexible rather than based on a strict progression from phase I to II to III; researchers could consider phase I studies with an expansion cohort, Phase I/II designs or phase II studies with a safety run in. Pharmacokinetics are recommended when interactions or overlapping toxicity is expected. Pharmacodynamic evaluations should be considered especially in a subset of patients closest to the recommended dose; an attempt should be made to validate surrogate tissues to enable inclusion for all patients. Schedule and or dose should be formally explored for e.g. with a randomised or an adaptive design. Data and knowledge sharing was strongly recommended, including the creation of formal or informal consortia of laboratories with individual expertise in pathway or target based models, collaboration between companies to ensure that agents which are ‘best in class’ are combined, and the development of databases which will be able to inform the development of future recommendations/guidelines.