Objective: Additional treatments with persisting benefit are needed for ADHD. Because ADHD often shows excessive theta electroencephalogram (EEG) power, low beta, and excessive theta-beta ratio (TBR), a promising treatment is neurofeedback (NF) downtraining TBR. Although several nonblind randomized clinical trials (RCTs) show a medium-large benefit for NF, a well-blinded, sham-controlled RCT is needed to differentiate specific from nonspecific effects. Method: Experts in NF, ADHD, clinical trials, and statistics collaborated to design a double-blind multisite RCT. Results/Conclusion: At four sites, 180 children aged 7 to 10 years with rigorously diagnosed ADHD and TBR greater than or equal to 5 will be randomized to active TBR-NF versus sham NF of equal duration, intensity, and appearance. Sham, utilizing prerecorded EEGs with participant artifacts superimposed, will keep participants and staff blind. Treatment fidelity will be trained/monitored by acknowledged NF leaders. Multidomain assessments before, during, and after treatment (follow-up to 2 years) will also include tests of blinding and sham inertness. (Contains 1 figure and 1 table.)