Serum insulin-like growth factor 1 (sIGF-1) is an important growth factor in childhood. However, studies on sIGF-1 among children from low-income countries are few, and the role of body composition is unknown. To assess the associations of anthropometry, body composition, inflammation, and breastfeeding with sIGF-1 among children with moderate acute malnutrition (MAM). A cross-sectional study based on admission data from 6- to 23-mo-old children with MAM participating in a nutrition intervention trial (Treatfood) in Burkina Faso. Linear regression analysis was used to identify correlates of sIGF-1. Among 1546 children, the median (IQR) sIGF-1 was 12 (8.2-18.3) ng/mL. sIGF-1 was highest at 6 mo, with a nadir ∼10-11 mo, and higher in girls than boys. Length-for-age z score (LAZ), weight-for-length z score (WLZ), and midupper arm circumference were positively associated with sIGF-1 (P≤0.001). Fat-free mass (FFM) was also positively associated, as sIGF-1 increased 1.5 (95% CI: 0.5, 2.5) ng/mL for each 1-kg increase in FFM. However, the association disappeared after adjustment for height. Elevated serum C-reactive protein and α1-acid glycoprotein were negatively associated with sIGF-1 (P≤0.001), as was fever (P<0.001) but not a positive malaria test per se (P=0.15). Children never breastfed had lower sIGF-1 (-5.1; 95% CI: -9.8, -0.3). LAZ and WLZ were positively and inflammation negatively associated with sIGF-1. As all children were moderately malnourished and many had inflammation, this probably explains the very low median sIGF-1. The association of FFM with sIGF-1 was fully explained by height. There was a marked age pattern, with a nadir in late infancy, confirming findings from smaller studies from well-nourished populations. There is a need for prospective studies to disentangle the role of sIGF-1 in growth and health. This trial was registered at https://www.isrctn.com as ISRCTN42569496.