Infant temperament and parental feeding practices may be risk factors for childhood obesity, however most studies have relied upon parent-report assessments. We tested whether infant emotional distress and maternal restrictive feeding at 12-months of age, assessed observationally at a home feeding interaction, predicted child BMI through age 6years. We conducted a prospective observational study of 86 children (34 girls and 52 boys, from 55 adoptive and 31 non-adoptive families) enrolled in the Colorado Adoption Project. Mother-infant feeding interactions were video-recorded during a home snack or meal at year 1, and child anthropometrics (length or height, and weight) were assessed at years 1 through 6. The main outcome measures were child weight-for-length at year 1 and body mass index (BMI: kg/m2) at years 2–6. Results of generalized linear models indicated that greater infant emotional distress at 12-months predicted greater increases in child weight status through age 6years, B=0.62 and odds ratio (OR)=1.87. In separate analyses, restrictive feeding interacted with child sex in predicting weight status trajectories (p=.012). Male infants whose mothers displayed any compared to no restriction at year 1 showed a downward BMI trajectory from 2 to 6years; for female infants, exposure to any compared to no restriction prompts predicted increasing BMI from 4 to 6years. In sum, early obesity prevention strategies should pay greater attention to infant temperament, especially distress and negative affect, and how parents respond to such cues. Additionally, ‘responsive feeding’ strategies that provide an alternative to restriction warrant greater research during infancy.