There has been an exponential growth in assessment methods to diagnose disorders on the autism spectrum. Many reasons for this trend exist and include advancing knowledge on how to make a diagnosis, the heterogeneity of the spectrum, the realization that different methods may be needed based on age and intellectual disability. Other factors include the recognition that some testing will be for diagnosis, while other diagnostic work will be for identifying targets for intervention and for monitoring symptoms over time. Measures of co-occurring conditions that interface with measures of core symptoms are also needed and should aid in multidisciplinary treatment. In addition to these factors we discuss factors associated with observational data collection methods, length, and scoring of evaluations and the fallacy that a "gold standard" for diagnosis of autism exists or in fact is desirable. The implications of these data are discussed.