OBJECTIVE: To examine the screening ability of the Strengths and Difficulties Questionnaire (SDQ) for ADHD subtypes in a clinical sample. METHOD: Parents of 523 children (3 to 17 years old) referred to Child and Adolescent Mental Health Services completed the Spanish version of the SDQ and the ADHD Rating Scale-IV. Receiver operating curve (ROC) curve analyses and likelihood ratios (LRs) were conducted. RESULTS: The LR results indicated that the 8/10 cutoff showed the highest diagnostic accuracy. The sensitivity of the SDQ 8/10 cutoff for the three subtypes was significantly different: 84.0% (95% confidence interval [CI] = [75.58, 89.90]) of ADHD combined subtype (ADHD-Co), 25.0% (95% CI = [17.55, 34.30]) of ADHD predominantly inattentive subtype (ADHD-I), and 77.8% (95% CI = [68.71, 84.83]) of ADHD predominantly hyperactive/impulsive subtype (ADHD-H). Sex and age differences were found. These screening differences were also found when using the 7/10 cutoff or the SDQ predictive algorithm. CONCLUSION: Our study supports the use of the SDQ in the screening for ADHD. However, not all ADHD subtypes are equally screened.