The natural history of facioscapulohumeral muscular dystrophy (FSHD) has not been studied prospectively. Knowledge of the natural progression of any disease provides essential information for the design of clinical trials. We present a protocol for the study of the natural history of FSHD using quantitative muscle testing (QMT), manual muscle testing (MMT), and functional testing. Thirty-two persons with FSHD (mean age = 36.1 years, SD = 9.6, range = 17-49) and 32 age- and gender-matched volunteer controls (mean age = 35.8 years, SD = 8.0, range = 23-50) served as subjects. Using standardized testing procedures, we examined intrarater reliability of the MMT, QMT, and functional testing measurements in both groups. We also examined interrater reliability in 7 subjects with FSHD. Eighteen muscle groups were tested for each subject using QMT and MMT. Intraclass correlation coefficient (ICC) values ranged from .86 to .99 for intrarater reliability and from .86 to .99 for interrater reliability of QMT measurements. Weighted kappa values of .81 to .98 for intrarater reliability and .50 to 1.00 for interrater reliability were obtained for MMT measurements. Intrarater ICCs for various functional testing measures ranged from .60 to .97. In addition, the comparability of the two QMT machines used in the study was demonstrated by testing the same set of volunteer controls on each machine's linear force transducer (ICC = .89-.98). We conclude that this standardized testing protocol produces reliable measurements of muscle strength and functional ability in subjects with FSHD.