Purpose: The Doha 2019 women’s World Championship marathon took place in extreme hot (32 °C), humid conditions (74% relative humidity) culminating in unprecedented (41%) failure rates. We explored whether extreme heat or suboptimal pacing was responsible for diminished performance against a temperate “control” (London 2017: 19 °C, 59% relative humidity) and whether physical characteristics (eg, body surface area, estimated maximal oxygen uptake, habitual heat exposure) explained performance. Method: Five-kilometer-pace (km·h⁻¹) data underwent repeated-measures analyses of hot (Doha, n = 40) versus temperate pacing and performance (London, n = 78) within and between marathon pacing (finisher quartiles normalized against personal best; n = 10 per group) and within hot marathon finishers versus nonfinishers (up to 10 km; normalized data). Possible predictors (multiple regression) of hot marathon pacing were explored. Tests to .05 alpha level, partial eta squared (${\eta }_{\mathrm{p}}^2$) indicates effect size. Results: Mean (SD) of Doha (14.82 [0.96] km·h⁻¹) pace was slower (London: 15.74 [0.96] km·h⁻¹; P = .00; ${\eta }_{\mathrm{p}}^2=.500$). In hot conditions, athletes finishing in positions 1 to 10 (group 1) started more conservatively (93.7% [2.1%] of personal best) than slower runners (groups 3 and 4; 96.6% [2.8%] of personal best; P < .05, ${\eta }_{\mathrm{p}}^2=.303$). Groups were not different at 15 km and then slowed immediately (groups 3 and 4) or after 20 km (group 2). Finishers and nonfinishers adopted similar pace up to 10 km (P > .05, ${\eta }_{\mathrm{p}}^2=.003$). World ranking predicted (P = .00; r² = .248) average pace in Doha. Conclusion: Extreme hot conditions reduced performance. Top 10 athletes adopted a conservative initial pace, whereas lower-placing athletes adopted a faster, aggressive start. Pacing alone does not explain high failure rates in nonfinishers. Athletes competing in the heat should initially pace conservatively to optimize performance.