Purpose: The initial diagnosis of stuttering is based on stuttering-like disfluencies. This approach fails to capture the variability in overt stuttering behaviors of persons who stutter across speaking tasks and over time. This preliminary study explored the relationships between day-to-day variability of stuttering and several commonly used self-report and behavioral outcomes used to diagnose and monitor stuttering in clinical and research settings. Method: Five adults who stutter participated in this study. All participants completed a demographic questionnaire; the Overall Assessment of the Experience of Stuttering, State, and Trait Anxiety Inventory; and the Brief Version of Unhelpful Thoughts and Beliefs About Stuttering (UTBAS), followed by six (i.e., twice weekly) online sessions where conversation and reading samples were video-recorded to determine stuttering frequency and severity. Participants also rated their stuttering severity, variability, stress, anxiety, and sleep on a 7-point Likert scale for each session. Results: No self-report measures correlated with stuttering frequency or severity measures determined at each twice-weekly online session. Positive correlations were observed between self-rated stuttering severity and stress, anxiety, and normal disfluencies. Stuttering variability across tasks and sessions was positively correlated with the one-time Overall Assessment of the Speaker's Experience of Stuttering (OASES) and UTBAS-6 scores. Conclusions: Stuttering frequency was found to vary to different degrees for individual participants, and preliminary data suggest that scores on standardized assessments such as the OASES and UTBAS-6 are predictive of stuttering variability. Objective data on stuttering frequency may not capture the entire experience of stuttering.