Purpose: This study aims to investigate the relationship between split disciplines and overall extreme ultra-triathlon (EUT) performance and verify the relationship among physiological and neuromuscular measurements with both fractional and total EUT performance while checking which variables could predict partial and overall EUT race time. Methods: Eleven volunteers (37 ± 6 years; 176.9 ± 6.1 cm; 77.9 ± 10.9 kg) performed two maximal graded tests (cycling and running) for physiological measurements and muscle strength/power tests to assess neuromuscular functions. Results: The correlation of swimming split times to predict overall EUT race times was lower than for cycling and running split times (r2 = 0.005; p > .05; r2 = 0.949; p < .001 and r2 = 0.925; p < .001, respectively). VO2peak obtained during running test (VO2peakrun) and VT power output assessed during cycling test (VTPO) were the highest predictors of cycling performance (r2 = 0.92; p = .017), whereas VO2peakrun and peakpower output in the cycling test (PPO) were the highest predictors of running performance (r2 = 0.94; p = .008). Conclusion: VO2peakrun and VTPO, associated to jump height assessed during countermovement jump (CMJ) test were the highest correlated variables to predict total EUT performance (r2 = 0.99; p = .007). In practical terms, coaches should include the assessment of VO2peakrun, VTPO, and CMJ to evaluate the athletes' status and monitor their performance throughout the season. Future studies should test how the improvement of these variables would affect EUT performance during official races.