The rate of success of novel treatments is low, especially in CNS diseases. The difficulty of demonstrating efficacy is especially difficult for disease-modifying (as opposed to symptomatic) therapies. Perceived difficulties limit the enthusiasm of pharmaceutical companies to undertake novel drug development. Most things fail, and the cost of each failure is prohibitive. Obstacles to CNS drug development include poor predictive value of preclinical (animal) models, incomplete understanding of disease-relevant biology, lack of reliable target engagement and pharmacodynamic biomarkers, imprecise clinical outcome measures, variability in the clinical population, and shortage of ‘trial-ready’ patients. Overcoming each of these obstacles can lower the perceived barrier to further drug development. Importantly, intermediate steps that inform on either target engagement or pharmacodynamic activity can provide crucial information as to the ultimate likelihood of success of a novel treatment.