A fundamental challenge to the timely diagnosis of Autism Spectrum Disorder (ASD) is the reliance on the observation of a set of aberrant behavior. Consequently, the diagnostic process requires that the child reach an age where the behaviors would typically be exhibited. The identification of a reliable biological marker (biomarker) could be of considerable benefit to the diagnostic process. As a diagnostic biomarker, porphyrins present an attractive prospect as previous studies have reported consistent findings of children with ASD showing significant elevations in porphyrin levels in contrast to controls. Furthermore, there is some evidence that ASD severity may be associated with porphyrins, which would be a valuable characteristic of any ASD biomarker. Importantly, for practical use, porphyrins can be tested non‐invasively via a sample of urine. The present study sought to investigate whether porphyrin profiles can reliably be used to (a) differentiate ASD cases from healthy controls; and (b) predict ASD severity. The study compared the porphyrin levels of three groups of children aged 2–6 years: Group 1—children diagnosed with ASD (n = 70); Group 2—healthy, normally developing siblings of children diagnosed with ASD (n = 36); and Group 3—healthy, normally developing children with no known blood relative diagnosed with ASD (n = 54). The results of logistic regression analyses failed to find support for the hypotheses that porphyrin levels could be used as a valid tool to detect ASD cases or predict severity. Autism Res 2014 , ●: ●●–●●. © 2014 International Society for Autism Research, Wiley Periodicals, Inc.