Abstract
<title>Abstract</title> <p>Background Ultramarathon participation has grown over 345% in the last decade, yet the psychological determinants of performance in these events remain quantitatively unsynthesized. Mental toughness (MT) and self-efficacy (SE) are the two constructs most consistently proposed as indicators of ultrarunning success, but primary findings conflict, ranging from null associations in an elite sample to small positive effects in broader field samples. No meta-analysis has pooled these effects. Methods Quantitative studies reporting an extractable association between MT or SE and an ultramarathon performance outcome (finish time, placing, performance index, or completion versus non-finish) were identified through electronic searches of openly indexed literature and systematic citation chasing of two prior reviews and all included studies. Effects were converted to a common correlation metric, transformed to Fisher's z, and pooled with a DerSimonian–Laird random-effects model. One study reporting only non-significant omnibus statistics without direction was conservatively imputed as a null effect and examined in sensitivity analysis. The review was not prospectively registered. Results Four studies met inclusion criteria (total N = 940). The pooled association between MT/SE and ultramarathon performance was small and positive, r = .20 (95% CI .13–.26, p < .001), with negligible heterogeneity (I² = 1%; 95% prediction interval .05–.33). Sensitivity analysis excluding the imputed null yielded r = .21 (95% CI .14–.27), and the completion-outcome subgroup yielded r = .20 (95% CI .12–.27). Conclusions MT and SE show a small, consistent positive association with ultramarathon performance and completion — approximately two-thirds the magnitude of the pre-event self-efficacy effect reported across sport generally. This attenuation is consistent with a threshold account in which high MT and SE are near-universal among those who reach an ultramarathon start line, restricting their range and limiting their discriminative value within the competing population.</p>