This study aimed to evaluate the reliability and the level of agreement of the ADR encoder to measure the mean propulsive velocity (MPV) of the bar in the bench press (BP) exercise on the Smith machine. Eleven males (21.6 ± 1.5 years; body mass 76.05 ± 9.73 kg) performed the protocol with isometric phase prior to concentric muscle action (PP) and the protocol in the absence of isometric phase (N-PP) for BP exercise on Smith machine. ADR encoder reported reliability values with almost perfect correlations in all training zones and protocols (PP: ICC = 0.940–0.999, r = 0.899–0.997, CV = 1.56%–4.05%, SEM = 0.0022–0.0153,and MDC = 0.006–0.031 m/s; N-PP: ICC = 0.963–0.999, r = 0.946–0.998, CV = 0.70%–3.01%, SEM = 0.0012–0.0099, and MDC = 0.003–0.027 m/s). Although the levels of agreement were high in both protocols (PP: SEM = 0.0024–0.0204 m/s, MDC = 0.007–0.057 m/s; N-PP: SEM = 0.0034–0.0288 m/s, MDC = 0.009–0.080 m/s), ADR encoder considerably underestimated the MPV values in both protocols (PP: t = −2.239 to −9.486, p < 0.001–0.01; N-PP: t = −6.901 to −17.871, p < 0.001) with respect to the gold standard (T-Force). In conclusion, ADR encoder offers high reliability for the measurement of MPV in bench press exercise performed on Smith machine regardless of their execution mode, in the entire range of intensities. However, this device is not interchangeable with T-Force since it considerably underestimates the MPV values, especially at low loads (0%–40%). Furthermore, the use of too wide load ranges suggests that the data be interpreted with caution, pending further research to corroborate the findings presented.