Loss of skeletal muscle is a frequent finding in patients with kidney disease. The identification of low muscle mass in the clinical practice is usually made by applying indirect methodologies, such as anthropometrics and bioelectrical impedance technologies, which have important limitations that hamper their use in the renal setting. In particular, measurements obtained with those two methods are prone to variability based on the fluid status of patients, which can be a problem in a population in which fluid overload is a frequent finding, a real problem in the renal setting. The aim of this dissertation was to develop a new, more accurate methodology for the assessment of skeletal muscle that can be used in routine clinical practice. On this regard, quadriceps muscle ultrasound was studied and applied in different settings of patients with kidney disease (acute and chronic settings). Findings of this research show that this new methodology, which is low cost and highly available in the clinical practice, is also accurate, sensitive and is not influenced by fluid status, allowing for a more precise assessment of muscle status of patients with kidney disease.