A major distinction among different skeletal muscles in the human body is the number, size, and arrangement of its cells, referred to as a muscle’s architecture. Muscle architecture is indicative of a muscle’s ability to contract and produce force and, like muscle function, is plastic. While neuromuscular plasticity is the basis of the field of physical medicine and rehabilitation, little is known about how the most functionally meaningful muscle architecture parameters (optimal fascicle length and physiological cross-sectional area (PCSA)) adapt in humans. Classically, muscle architecture has been quantified in cadaver dissections;​ functional parameters are computed from measurements at the whole muscle, fascicle, and sarcomere scales. However, a muscle’s ability to adapt to altered stimulus, use, injury, or disease, cannot be investigated in non-living tissue. In vivo imaging is increasingly adopted to capture muscle structure in living tissue. However, uncertainty remains regarding the validity and reliability of quantitative measures obtained using these methods, which affects the confidence with which muscle adaptation can be accurately studied in humans. In addition, in vivo muscle architecture is rarely measured across all of the critical scales needed to compute optimal fascicle length and PCSA. My dissertation establishes confidence in the ability to reliably quantify multiscale in vivo muscle structure and demonstrates the extent to which current approaches can detect structural adaptation in human muscle. To accomplish this, I first demonstrate the reliability of two imaging techniques (extended field-of-view ultrasound and second harmonic generation microendoscopy) which enable measurement at different scales (fascicle length and sarcomere length, respectively). I then combine these and other multiscale imaging techniques to establish interlimb variability in muscle-tendon architecture of multiple primary wrist muscles, in vivo, in adults without impairments. Finally, I characterize the magnitude of multiscale muscle architectural adaptation observed in chronic muscle shortening in two populations with unilateral impairment (including multiple wrist muscles of an individual with an orthopedic surgical repair and the biceps brachii in a population with neurological impairment). Most notably, I show that in vivo architectural differences between limbs following unilateral impairment are substantially larger than both the uncertainty from in vivo imaging measures and natural interlimb variability in participants without impairments. The ability to quantify in vivo muscle adaptation reliably with minimally invasive in vivo imaging techniques ultimately enables the development of more targeted rehabilitation and surgical interventions for individuals with maladapted muscle.