OBJECTIVES: To elicit caregivers’ preferences for evidence-based treatment options for their child’s attention deficit hyperactivity disorder (ADHD), and to identify segments of caregivers who display similar preferences. METHODS: Caregivers with a child aged 4–14 and in care for ADHD were recruited from outpatient clinics and advocacy groups. All caregivers completed a self-administered survey that included socio-demographic information, and a best-worst scaling (BWS) instrument assessing treatment preferences. The BWS instrument comprised 18 choice tasks, each displaying seven treatment attributes: medication, therapy, school involvement, caregiver behavior training, physician management, provider communication and out-of-pocket costs. Every attribute was operationalized into 3 possible levels. Within each task, caregivers selected one best and one worst attribute. A scale-adjusted latent-class (SALC) analysis was conducted to account for variability in the consistency of responses. RESULTS: Our study population of 164 caregivers were on average 42 years old (SD 8.7), predominantly female (95%), white (65%), married (61%), college-educated (73%), and 20% had a child who was diagnosed with ADHD for ≤1year. Based on the aggregate results, using medication everyday was the most preferred treatment attribute (coefficient=2.41, p<0.001). Three latent classes (i.e. segments) that best described the data were identified, and the scale factor included in the model was significant (p<0.001). The 3 segments comprised 28%, 27%, and 45% of our study population. Segment 1 has the strongest preference for ‘medication’ (coefficients=3.69 –4.34, all p<0.001) while Segment 2 displayed the least preference for medication (coefficients= -1.49 – -3.36, all p<0.001).  Segment 3 was most cost-avoidant (coefficients=-2.13 – -6.11, all p<0.001) but had the strongest preference for ‘school involvement’ (coefficients=0.63 – 2.58, all p<0.05). CONCLUSIONS: This study demonstrated variation in caregivers’ priorities for ADHD treatment attributes.  A better understanding of preferences for evidence-based treatment options can enhance patient-centered care. By utilizing SALC, our study reduces the likelihood of misclassification error.