Attention-deficit/hyperactivity disorder (ADHD) is common, chronic, and associated with significant functional impairment. It is highly treatable. It is therefore not only a major public health problem but also one that provides a unique opportunity in medicine to make a significant difference. This article will discuss the methodology needed to demonstrate empirically the impact of treatment on actual burden of illness in practice. Where efficacy studies demonstrate whether a treatment can work, effectiveness studies tell us whether they actually do work. Clinical trials exclude incompetent, non-compliant, and seriously comorbid patients, so that the information obtained from these trials tells us the most about the patients we see the least. Small differences in effect size in pivotal trials of efficacy have become a key variable for rating treatments as first line or second line, without consideration of effectiveness variables such as comorbidity, difficulty with appetite or sleep, patient preference, capacity for compliance, timing of functional impairment, and substance use. These effectiveness variables are less well studied, but critical to clinical decision making. In reality, fewer than 10% of our patients comply with and persist with treatment. To learn more about why patients are discontinuing treatment, we need to explore measures of effectiveness empirically. Effectiveness studies are also important to provide regulatory bodies with the data they need to balance the risk of treatment with the risk of failing to treat. Practical clinical trials and naturalistic follow-up studies will allow us to evaluate the true clinical impact of short-term efficacy trials.