Dietary patterns describe the combination of foods and beverages in a diet and the frequency of habitual consumption. Better understanding of childhood dietary patterns and antenatal influences could inform intervention strategies to prevent childhood obesity. We derived empirical dietary patterns in 1142 children (average age 6·0 (sd 0·2) years) in New Zealand, whose mothers had participated in the Screening for Pregnancy Endpoints (SCOPE) cohort study and explored associations with measures of body composition. Participants (Children of SCOPE) had their diet assessed by FFQ, and dietary patterns were extracted using factor analysis. Three distinct dietary patterns were identified: ‘Healthy’, ‘Traditional’ and ‘Junk’. Associations between dietary patterns and measures of childhood body composition (waist, hip, arm circumferences, BMI, bioelectrical impedance analysis-derived body fat % and sum of skinfold thicknesses (SST)) were assessed by linear regression, with adjustment for maternal influences. Children who had higher ‘Junk’ dietary pattern scores had 0·24 (sd 0·08; 95 % CI 0·04, 0·13) cm greater arm and 0·44 (sd 0·05; 95 % CI 0·01, 0·10) cm greater hip circumferences and 1·13 (sd 0·07; 95 % CI 0·03, 0·12) cm greater SST and were more likely to be obese (OR 1·74; 95 % CI 1·07, 2·82); those with higher ‘Healthy’ pattern scores were less likely to be obese (OR 0·62; 95 % CI 0·39, 1·00). In a large mother–child cohort, a dietary pattern characterised by high-sugar and -fat foods was associated with greater adiposity and obesity risk in children aged 6 years, while a ‘Healthy’ dietary pattern offered some protection against obesity. Targeting unhealthy dietary patterns could inform public health strategies to reduce the prevalence of childhood obesity.