The detection of a secondary inorganic phosphate (Pi) resonance, a possible marker of mitochondrial content in vivo, using phosphorus magnetic resonance spectroscopy (31P-MRS), poses technical challenges at 3Tesla (T). Overcoming these challenges is imperative for the integration of this biomarker into clinical research. To evaluate the repeatability and reliability of measuring resting skeletal muscle alkaline Pi (Pialk) using with 31P-MRS at 3 T. After an initial set of experiments on five subjects to optimize the sequence, resting 31P-MRS of the quadriceps muscles were acquired on two visits (~4days apart) using an intra-subjects design, from 13 sedentary to moderately active young male and female adults (22 ± 3years old) within a whole-body 3T MR system. Measurement variability attributed to changes in coil position, shimming procedure, and spectral analysis were quantified. 31P-MRS data were acquired with a 31P/-proton (1H) dual-tuned surface coil positioned on the quadriceps using a pulse-acquire sequence. Test-retest absolute and relative repeatability was analyzed using the coefficient of variation (CV) and intra-class correlation coefficients (ICC), respectively. After sequence parameter optimization, Pialk demonstrated high intra-subject repeatability (CV: 10.6 ± 5.4%, ICC: 0.80). Proximo-distal change in coil position along the length of the quadriceps introduced Pialk quantitation variability (CV: 28 ± 5%), due to magnetic field inhomogeneity with more distal coil locations. In contrast, Pialk measurement variability due to repeated shims from the same muscle volume (0.40 ± 0.09mM; CV: 6.6%), and automated spectral processing (0.37 ± 0.01mM; CV: 2.3%), was minor. The quantification of Pialk in skeletal muscle via surface coil 31P-MRS at 3T demonstrated excellent reproducibility. However, caution is advised against placing the coil at the distal part of the quadriceps to mitigate shimming inhomogeneity.