Background Undernutrition contributes to at least half the estimated six million annual childhood deaths worldwide. Furthermore, one in three children fails to meet their developmental potential because of risks including stunting, illness, under‐stimulation, poor responsive interactions and maternal depressive symptoms. Our study investigates the role of caregiving processes on children's height‐for‐age at 2 and 4 years. Methods The Pakistan Early Child Development Scale‐up study assessed the longitudinal effectiveness of early nutrition and responsive stimulation interventions on growth and development at 4 years of age. In total, 1302 children were followed up from birth to 4 years. We leveraged path analyses to explore potential mediators of early intervention effects on children's height‐for‐age at 4 years, including maternal depressive symptoms, mother–child interaction quality, diarrhoeal illness and height‐for‐age at 2 years. Results Our final model had excellent model fit (comparative fix index = 0.999, Tucker–Lewis index = 0.998, root mean square error of approximation = 0.008) and showed that mother–child interaction quality mediated the effects of both enhanced nutrition and responsive stimulation interventions on height‐for‐age at 4 years via its longitudinal stability from 2 years of age (β = 0.016, p = 0.005; β = 0.048, p < 0.001, respectively). Further, diarrhoeal illness mediated the effects of maternal depressive symptoms at 1 year post partum on children's height‐for‐age at 4 years via the longitudinal stability of height‐for‐age z‐score from 2 years of age onwards (β = −0.007, p = 0.019). Conclusions The quality of early caregiving experience mediated the association between both interventions and height‐for‐age. The effect of maternal depressive symptoms on growth was mediated by diarrhoeal illness. Programmatic approaches to child nutrition and growth must address all these potentially modifiable factors.