We explored the influence of credibility and evidence on public perceptions of ASD treatments using survey methodology. Participants (N = 379) read texts about different ASD treatments. The text presentation was based on a 2 × 2 within-subjects factorial design with treatment status [evidence based practices (EBP) vs. non-EBP] and source credibility in the text (credible vs. non-credible) as the independent variables. An instructional manipulation condition served as a between subjects factor. Respondents were more familiar with non-EBPs than EBPs, but viewed EBPs as being more credible and were more likely to endorse them compared to pseudoscientific practices. Interactions between source credibility and instructional manipulation were found on ratings of credibility and recommendation of both EBP and non-EBP texts. Implications of these findings are discussed.