This article describes the initial development and testing of the Antiretroviral Medication Complexity Index (AMCI), an instrument to measure complexity of antiretroviral medications and regimens. Data were derived from a pilot study and the main study of an intervention designed to promote adherence to antiretroviral medications. Interrater reliability coefficients ranged from .74 to 1.0 for project staff and from .33 to 1.0 for nurse clinicians. Test-retest coefficients ranged from .297 to .988 for forms completed by the same interviewer and from .200 to .922 for different interviewers. Strong relationships existed between rankings of the medications by clinical judgment and rankings of the medications using the AMCI, providing support for criterion validity. Complexity as measured by the AMCI was associated with some measures of adherence, providing partial support for construct validity. Results provide evidence for the reliability and validity of the AMCI when used for research purposes.