Our aims were to measure anthropometric and oxygen uptake (V˙O2) variables in the laboratory, to measure kinetic and stride characteristics during a trail running time trial, and then analyse the data for correlations with trail running performance. Runners (13 men, 4 women: mean age: 29 ± 5 years; stature: 179.5 ± 0.8 cm; body mass: 69.1 ± 7.4 kg) performed laboratory tests to determine V˙O2 max, running economy (RE), and anthropometric characteristics. On a separate day they performed an outdoor trail running time trial (two 3.5 km laps, total climb: 486 m) while we collected kinetic and time data. Comparing lap 2 with lap 1 (19:40 ± 1:57 min vs. 21:08 ± 2:09 min, P < 0.001), runners lost most time on the uphill sections and least on technical downhills (-2.5 ± 9.1 s). Inter-individual performance varied most for the downhills (CV > 25%) and least on flat terrain (CV < 10%). Overall stride cycle and ground contact time (GCT) were shorter in downhill than uphill sections (0.64 ± 0.03 vs. 0.84 ± 0.09 s; 0.26 ± 0.03 vs. 0.46 ± 0.90 s, both P < 0.001). Force impulse was greatest on uphill (248 ± 46 vs. 175 ± 24 Ns, P < 0.001) and related to GCT (r = 0.904, P < 0.001). Peak force was greater during downhill than during uphill running (1106 ± 135 vs. 959 ± 104 N, P < 0.01). Performance was related to absolute and relative V˙O2 max (P < 0.01), vertical uphill treadmill speed (P < 0.001) and fat percent (P < 0.01). Running uphill involved the greatest impulse per step due to longer GCT while downhill running generated the highest peak forces. V˙O2 max, vertical running speed and fat percent are important predictors for trail running performance. Performance between runners varied the most on downhills throughout the course, while pacing resembled a reversed J pattern. Future studies should focus on longer competition distances to verify these findings and with application of measures of 3D kinematics.