This study was undertaken in Afar Regional state, Ethiopia with the aim of developing a body weight prediction model from linear measurements. Stratified random sampling was employed to select the study area and households. A total of 891 goats (823 female and 68 male) above six months of age were used for body measurements. Each measurement was sexed and aged using dentition. Body measurements were analyzed using the Generalized Linear Model of the Statistical computing R. Tukey’s HSD post-hoc test was used to separate means if analysis of variance showed significance. Pearson’s correlation coefficients were used to select variables that have a strong correlation with body weight. Then stepwise regression procedure of R was employed to identify a model. The overall mean body weight was 23.8 kg. Body measurements for males and females were (26.44 and 20.87 kg) for body weight, (64.8 and 59.5 cm) for chest girth, (63.1 and 58.9 cm) for body length, (59.6 and 55.4 cm) for height at wither and (13.3 and 12.4 cm) for pelvic width respectively. In all measurements, males were significantly (P<0.01, P<0.05) heavier than females. Most body measurements increased consistently as age advanced. Most of the linear parameters depicted positive and highly significant (P<0.01) correlation with body weight. The body weight of the Afar goat breed could be predicted with higher accuracy from chest girth (CG) and body length (BL) in pastoral areas where there is limited weighing scale to improve the breed, husbandry practice and weight based marketing.