Body condition scoring (BCS) is a subjective semiquantitative method of assessing body fat and muscle. Scoring systems use a scale in which the midrange represents optimal body condition, lower values represent lean to emaciated conditions, and higher values indicate excessive body fat. A valid BCS system is clearly described, relevant to the species, shows agreement within and between raters, and is consistent with objective measures. The goal of the current study was to assess intra- and interrater variability of a BCS system that uses a 1-to-5 scale and entails the palpation of key anatomic sites (hips, spine, pelvis, thorax, and abdomen) to assess prominence of bony structures, muscle mass, and subcutaneous fat. To assess interrater variability, 4 raters independently assessed BCS in 616 rhesus macaques (Macaca mulatta) in 4 age groups: infant, younger than 1 y; juvenile, 1 to 4 y; subadult, 4 to 7 y; and adult, 7 to 17 y. To assess intrarater variability, each rater independently reevaluated a subset of adult macaques (n = 15) within 2 wk of initial evaluation. A weighted κ score was used to analyze intra- and interrater variability. Agreement between raters was highest for subadult and adult macaques, intermediate for juveniles, and least for infants. Intrarater agreement was high for all raters except one, for which it was moderate. Our results suggest that raters applied the BCS system most consistently to adult and subadult macaques and less so to juvenile and infant animals. However, the percentage agreement between raters to within one half of a score unit increased markedly when raters scored infants in the context of 'as is' rather than 'ideal for age.'