Resumo. ABSTRACT Machines and free weights are commonly used to increase muscle strength and performance. However, the use of elastic devices needs scientific investigations regarding its agreement against gold standards methods muscle strength. Aim: To determine the agreement of the 10-RM test between weight machine and elastic resistance in upper and lower limb exercises, and the intra-device reliability, agreement, and minimal detectable changes of the 10-RM test. Methods: Twenty-three healthy and untrained adults, of both sexes (age 22.43 ± 2.55), performed the 10-RM bench press and knee extension test with weight machines and elastic tubes, randomly. Results: The comparison of the 10-RM test between instruments, by the use of ICC index, demonstrated very high and high reliability for lower (ICC = 0.87) and upper (ICC = 0.80) limbs, respectively. The Bland Altman agreement analysis showed that the 10-RM test using elastic resistance produced lower loads (Kgf) on knee extension (34.16 ± 6.81 [elastic]; 36.22 ± 9.45 [machine]; Mean difference = 2.05 kgf ± 5.32) and higher loads (Kgf) on the vertical bench press (31.61 ± 7.81 [elastic]; 24.96 ± 10.11 [machine]; Mean difference = 6.65 kgf ± 4.93) when compared to the weight machines. Intra-device reliability also demonstrated a very high ICC index for upper and lower limbs on both devices. Conclusion: The use of 10-RM with elastic resistance presents high reliability and agreement when compared to weight machines for bench press or knee extension exercises.