Achilles tendon strain can be quantified using dynamic ultrasound, but its use in running is limited. Minimal effects on running pattern and acceptable test–retest reliability of muscle–tendon junction (MTJ) tracking are prerequisites for ultrasound use during running. We aimed to assess (i) the effect of wearing an ultrasound transducer on running pattern and (ii) the test–retest reliability of MTJ tracking during running. Sixteen long-distance runners (nine injury-free, seven with Achilles tendinopathy) ran at different speeds on an instrumented treadmill with a 10-camera system tracking skin-mounted retroreflective markers, first without and then with an ultrasound transducer attached to the lower leg to track the MTJ of the gastrocnemius medialis. Spatiotemporal parameters, joint kinematics and kinetics were compared between conditions using mixed ANOVAs and paired t-tests. MTJ tracking was performed manually twice by three raters in ten participants. Variability and standard error of measurement (SEM) quantified the inter- and intra-tester test–retest reliability. The running pattern was not affected by wearing the ultrasound transducer, except for significantly less knee flexion during midstance (1.6°) and midswing (2.9°) found when wearing the transducer. Inter-rater and intra-rater SEMs for MTJ tracking to assess the tendon strain (0.43%, and 0.56%, respectively) were about four times as low as between-group differences presented in literature. The minimal effects found on the running pattern and acceptable test–retest reliability indicates that dynamic ultrasound during running can be appropriately used to study Achilles tendon mechanics and thereby help improve our understanding of Achilles tendon behavior during running, injury development and recovery.