Background: A recent randomized controlled trial (N = 140) was indicative of large and sustained average improvements of Internet-based exposure for fibromyalgia, as compared to a waitlist. However, little is known about who benefits the most from this treatment. Objectives: To test for potential moderating effects of age, educational attainment, the duration of fibromyalgia, baseline overall fibromyalgia severity, pain intensity, fibromyalgia-related avoidance behaviour and symptom preoccupation on the waitlist-controlled effect of 10 weeks of Internet-based exposure for fibromyalgia. Methods: Secondary analysis of a randomized controlled trial (ClinicalTrials.gov NCT02638636). We used linear mixed effects models to determine whether the waitlist-controlled effect of exposure therapy on overall fibromyalgia severity (Fibromyalgia Impact Questionnaire) differed as a function of the potential moderators. Results: Only pain intensity (0–10) was found to be a significant moderator, where a higher baseline pain intensity predicted a more limited waitlist-controlled effect of Internet-based exposure (B = 3.48, 95% CI: 0.84–6.13). Standardized point estimates of effects were small for the sociodemographic variables, and in the moderate range for some clinical variables that did not reach statistical significance such as behavioural avoidance and time with the fibromyalgia diagnosis. Conclusions: Results suggest that Internet-based exposure treatment was more useful for participants with lower baseline levels of pain, and less so for participants with higher baseline levels of pain. The treatment had relatively similar effects across the other tested moderators. Significance: This study evaluated potential effect moderators in exposure-based treatment for fibromyalgia. Results indicated that a higher level of pain intensity at baseline was predictive of a less favourable outcome. It may be extra important to monitor progression for these patients and to provide additional therapeutic support when needed.