Background Lipid-based nutrient supplements (LNS) and corn–soy blends (CSBs) with varying soy and milk content are used in treatment of moderate acute malnutrition (MAM). We assessed the impact of these supplements on child development. Methods and findings We conducted a randomised 2 × 2 × 3 factorial trial to assess the effectiveness of 12 weeks’ supplementation with LNS or CSB, with either soy isolate or dehulled soy, and either 0%, 20%, or 50% of protein from milk, on child development among 6–23-month-old children with MAM. Recruitment took place at 5 health centres in Province du Passoré, Burkina Faso between September 2013 and August 2014. The study was fully blinded with respect to soy quality and milk content, while study participants were not blinded with respect to matrix. This analysis presents secondary trial outcomes: Gross motor, fine motor, and language development were assessed using the Malawi Development Assessment Tool (MDAT). Of 1,609 children enrolled, 54.7% were girls, and median age was 11.3 months (interquartile range [IQR] 8.2–16.0). Twelve weeks follow-up was completed by 1,548 (96.2%), and 24 weeks follow-up was completed by 1,503 (93.4%); follow-up was similar between randomised groups. During the study, 4 children died, and 102 children developed severe acute malnutrition (SAM). There was no difference in adverse events between randomised groups. At 12 weeks, the mean MDAT z-scores in the whole cohort had increased by 0.33 (95% CI: 0.28, 0.37), p < 0.001 for gross motor; 0.26 (0.20, 0.31), p < 0.001 for fine motor; and 0.14 (0.09, 0.20), p < 0.001 for language development. Children had larger improvement in language z-scores if receiving supplements with milk (20%: 0.09 [−0.01, 0.19], p = 0.08 and 50%: 0.11 [0.01, 0.21], p = 0.02), although the difference only reached statistical significance for 50% milk. Post hoc analyses suggested that this effect was specific to boys (interaction p = 0.02). The fine motor z-scores were also improved in children receiving milk, but only when 20% milk was added to CSB (0.18 [0.03, 0.33], p = 0.02). Soy isolate over dehulled soy increased language z-scores by 0.07 (−0.01, 0.15), p = 0.10, although not statistically significant. Post hoc analyses suggested that LNS benefited gross motor development among boys more than did CSB (interaction p = 0.04). Differences between supplement groups did not persist at 24 weeks, but MDAT z-scores continued to increase post-supplementation. The lack of an unsupplemented control group limits us from determining the overall effects of nutritional supplementation for children with MAM. Conclusions In this study, we found that child development improved during and after supplementation for treatment of MAM. Milk protein was beneficial for language and fine motor development, while suggested benefits related to soy quality and supplement matrix merit further investigation. Supplement-specific effects were not found post-intervention, but z-scores continued to improve, suggesting a sustained overall effect of supplementation. Trial registration ISRCTN42569496