To assess the validity and reliability of the Garmin Vector against the SRM power meter, 6 cyclists completed 3 continuous trials at power outputs from 100-300 W at 50-90 rev·min and a 5-min time trial in laboratory and field conditions. In field conditions only, a 30-s sprint was performed. Data were compared with paired samples t-tests, with the 95% limits of agreement (LoA) and the typical error. Reliability was calculated as the coefficient of variation (CV). There was no significant difference between the devices in power output in laboratory (p=0.245) and field conditions (p=0.312). 1-s peak power was significantly different between the devices (p=0.043). The LoA were ~1.0±5.0 W and ~0.5±0.5 rev·min in both conditions. The LoA during the 30-s sprint was 6.3±38.9 W and for 1-s peak power it was 18.8±17.1 W. The typical error for power output was 2.9%, while during sprint cycling it was 7.4% for 30-s and 2.7% for 1-s peak power. For cadence, the typical error was below 1.0%. The mean CVs were ~1.0% and ~3.0% for the SRM and Garmin, respectively. These findings suggest, that the Garmin Vector is a valid alternative for training. However, during sprint cycling there is lower agreement with the SRM power meter. Both devices provide good reliability (CV<3.0%).