Professional diagnosis would involve a standardized set of questions, like ADI-R http://www.wpspublish.com/store/p/2645/autism-diagnostic-interview-revised-adi-

Professional diagnosis would involve a standardized set of questions, like ADI-R http://www.wpspublish.com/store/p/2645/autism-diagnostic-interview-revised-adi-r http://www.wpspublish.com/store/p/2645/autism-diagnostic-int... or ADOS module 4 https://en.wikipedia.org/wiki/Autism_Diagnostic_Observation_Schedule https://en.wikipedia.org/wiki/Autism_Diagnostic_Observation_... Often the clinician would also want to interview parents, spouses, friends to get a third-party perspective on a person's behavior or ask about their behavior as a child. It is a relatively objective process - it does screen out 99% of people, and clinicians usually (80%-ish of the time IIRC) get the same yes-or-no answer for the same person. There's no medical test because (like most of the DSM) autism is a description, rather than an underlying cause. That is, if you diagnose flu using a set of symptoms, you're using the symptoms to predict that the virus exists. If you diagnose autism using the checklist in the DSM, you're simply saying that you saw the stuff on the list - there's nothing said about cause. See also here: http://intellectualizing.net/2015/03/31/not-explaining-autism/ http://intellectualizing.net/2015/03/31/not-explaining-autis... This is a fundamental difference between DSM diagnosis (which is putting people in descriptive categories, without explaining "why") and medical diagnosis of many conditions with a known cause (which means identifying an explanation such as a virus).