Diagnosing adult ADHD is frequently problematic because behavioral information from the patient's childhood, and multiple informants who can delineate the patient's current behavior, are often unavailable. This preliminary study was designed to explore whether objective neuropsychological testing may be a useful adjunct in the diagnosis of adult ADHD. Nineteen adults diagnosed with ADHD according to DSM-IV criteria, along with 10 controls, were assessed using a neuropsychological battery which comprised tests assessing linguistic, visual-spatial perceptual, academic, attentional and inhibitory control, mnestic and executive functions. Following preliminary analyses, designed to determine which variables best discriminated the groups, receiver operating characteristic (ROC) curves were constructed to determine the sensitivity and specificity of the best measures both alone and in combination. Only three measures significantly (p < 0.01) distinguished the groups; Digits Backwards from the WAIS-R and two reaction time measures from a computerized task modeled after Luria's Competing Motor Programs. ROC curve analyses indicated that in combination these measures had greater than 90% accuracy for classifying ADHD and non-ADHD patients. While further research is necessary these preliminary findings suggest that neuropsychological testing may be a useful adjunct in the differential diagnosis of adult ADHD.