Objective: The cognitive characterization of Alzheimer’s disease risk states, such as amnestic mild cognitive impairment (aMCI) and subjective cognitive decline (SCD), is fundamental for timely diagnosis and interventions. The Face Name Associative Memory Exam (FNAME) is sensitive to early Alzheimer’s disease brain changes, and an extended version captures a fuller range of associative memory abilities. We aimed to assess group effects in the extended FNAME in older adults with SCD, aMCI, and older adult controls (CON). Method: Two concurrently created versions of the extended FNAME were used to test three groups of older adults (CON = 35, SCD = 37, aMCI = 31) at two sites (Mexico = 59, Netherlands = 44). Extended FNAME memory abilities were analyzed in five analyses of variance. Group and site were considered as independent variables. For the recall ability, subtest levels were entered as a within-subject variable. The remaining abilities (Face Recognition, Name Recognition, Spontaneous Name Recall, and Face–Name Matching) were analyzed in independent models. Results: In all models, the main effect for group was significant with large effect sizes, driven by a worse performance of aMCI participants. No significant differences were found between SCD and CON. The main effect for site was only significant in Face Recognition. Conclusions: The worse performance of aMCI in the extended FNAME implies an impairment in associative memory abilities beyond recall. The similar performance of CON and SCD might be explained by the recruitment of SCD participants that did not spontaneously seek help for memory decline.