OBJECTIVES: In 2013, the Amyloid Imaging Taskforce (AIT) proposed that individuals with unexplained mild-cognitive impairment, possible AD, and early-onset dementia are potentially appropriate for amyloid PET in the diagnosis of AD and other forms of dementia. This analysis quantified the budgetary impact of using amyloid PET according to the AIT’s criteria from the US payer perspective. METHODS: An Excel-based model was developed for this analysis. The model projects the number of patients eligible for amyloid PET over a 3-year time horizon and calculates the incremental cost of using amyloid PET by considering direct medical resource uses for diagnostic work-up, dementia medications, and dementia care for patients staying in communities or institutions. Amyloid PET’s cost could be potentially offset by reducing diagnostic work-up time and resources use, unnecessary dementia  treatments and delaying disease progression and time to institutional care. The model was mainly populated with data from claims data analysis (Truven MarketScan®  2007–2012) and a survey of 75 dementia practitioners in the US, supplemented with data from literature, public databases, and assumptions. All costs were estimated in 2013 dollars.    RESULTS: Assuming an uptake of 5% incrementally each year, using amyloid PET increased the total cost by about $560,000 to $1,140,000 per 1 million covered lives over 3 years (or $0.016 to $0.032 per member per month) when amyloid PET’s cost was set between $2,500 and $6,000 per scan. Amyloid PET’s cost was partially offset by reductions in diagnostic work-up time and resources use and costs of institutional care due to more timely and accurate diagnosis and treatment. These results were most sensitive to variations in the cost of amyloid PET and time horizon.    CONCLUSIONS: Using amyloid PET in the diagnosis of AD according to the AIT’s criteria could result in better clinical outcomes with little impact on the annual budget.