Low nutritional value of complementary foods is associated with high incidence of childhood growth stunting in low-income countries. This study was done to test a hypothesis that dietary complementation with lipid-based nutrient supplements (LNS) promotes linear growth and reduces the incidence of severe stunting among at-risk infants. A total of 840 6-month-old healthy infants in rural Malawi were enrolled to a randomised assessor-blinded trial. The participants received 12-month supplementation with nothing, milk–LNS, soy–LNS, or corn–soy blend (CSB). Supplements provided micronutrients and approximately 280 kcal energy per day. Outcomes were incidence of severe and very severe stunting [length-for-age z-score, (LAZ) < −3.00 and <−3.50, respectively], and change in LAZ. The incidence of severe stunting was 11.8%, 8.2%, 9.1% and 15.5% (P = 0.098) and that of very severe stunting 7.4%, 2.9%, 8.0% and 6.4% (P = 0.138) in control, milk–LNS, soy–LNS and CSB groups, respectively. Between 9 and 12 months of age, the mean change in LAZ was −0.15, −0.02, −0.12 and −0.18 (P = 0.045) for control, milk–LNS, soy–LNS and CSB groups, respectively. There was no significant between-group difference in linear growth during other age-intervals. Although participants who received milk–LNS had the lowest incidence of severe and very severe stunting, the differences between the groups were smaller than expected. Thus, the results do not provide conclusive evidence on a causal association between the LNS supplementation and the lower incidence of stunting. Exploratory analyses suggest that provision of milk–LNS, but not soy–LNS promotes linear growth among at-risk infants mainly between 9 and 12 months of age.