Purpose: This study was designed to answer three questions: (a) Does percentage of syllables stuttered (%SS) differ between standard and challenge phone calls; (b) Does anxiety differ between standard and challenge phone calls; and (c) Is there a relationship between %SS and anxiety during standard and challenge phone calls? Method: Participants were 230 adults diagnosed with stuttering, who were participants from five clinical trials. Each participant received two 10-min phone calls at pretreatment and a further two phone calls 6 months or 20 weeks postrandomization. One phone call was standard, and the other presented challenge: occasionally disagreeing with, interrupting, and talking over participants, or asking for clarification of their views. Results: Statistically significant, but clinically minor, increases of %SS and anxiety occurred during the challenge phone calls. There was a statistically significant association between %SS and anxiety. Conclusions: Variable phone call procedures to assess stuttering severity in clinical trials are not likely to spuriously inflate or deflate treatment outcomes to a clinically important extent. Regardless, the present results suggest that there is statistical merit in controlling the nature of phone calls during clinical trials with the simple and replicable method developed in this report. Additionally, there is procedural merit in the challenge phone call procedure; it is a more valid representation of the challenges of everyday speech than the standard procedure. However, a disadvantage of the challenge phone call procedure is the practical issues associated with its use. The clinical and theoretical applications of the results are discussed.