Background and aims: We aimed to investigate the test-retest reliability and validity of ultrasound for two commonly used types of transducer, using different methods for the estimation of muscle size and echo intensity (EI). Methods: Fourteen healthy adults were included in this study. Ultrasound images of the rectus femoris size (thickness in cm and cross-sectional area [CSA] in cm2), obtained at the mid-thigh, were validated against MRI. Both a linear and a curved array transducer were used to assess rectus femoris size and EI (values 0–255, higher scores indicating increased intramuscular fat and interstitial fibrous tissue). To assess test-retest reliability of ultrasound, participants were tested twice, with a one-week interval. Validity and reliability were evaluated using paired sample t-tests, intraclass correlation coefficient (ICC), and Bland–Altman plots. Results: No significant differences between the repeated evaluations of rectus femoris thickness, CSA and EI were found. Reliability for thickness and CSA evaluations was excellent for both transducers (ICC = 0.87–0.97) and moderate for EI (ICC = 0.42–0.44). Mean difference between MRI and ultrasound for CSA (curved = 0.59 cm2, p = 0.086; linear = 2.1 cm2, p = 0.002) and thickness (curved = 0.31 cm, p = 0.01; linear = 0.21 cm, p = 0.043) were small but significant, except for CSA using a curved transducer. Agreement between ultrasound and MRI ranged from moderate for thickness (ICC = 0.45) to excellent for CSA (ICC = 0.92). Conclusions: Our study demonstrates that the test-retest reliability and validity of muscle size estimation by ultrasound for both curved and linear array transducers seems to be adequate. Future studies should focus on the longitudinal evaluation of muscle size and EI by ultrasound.