summary The dentition of 57 children aged 6–8 years old was examined clinically and radiographically. Subjects were divided into groups according to state of the dentition: caries free, no pathology and no pain; caries, no pathology and no pain; caries, pathology and no pain; caries, pathology and pain. The existence of a preferred chewing side (PCS) was determined using a visual spot-checking method. Subjects were instructed to chew gum on their posterior teeth. After 15 s the position of the chewing gum was recorded as right or left. This procedure was carried out seven times consecutively. Subjects had an ‘observed preferred chewing side’ (OPCS) when they chewed 5/7, 6/7 or 7/7 times on the same side. To determine if children were aware of their chewing side preference each subject's stated PCS was then recorded as left, right or none. ‘Coincidence’ between stated and OPCSs was said to occur if the stated PCS corresponded to the OPCS. Seventy-seven per cent of children had a PCS. The majority (70%) of caries free children also had a PCS (P < 0·05). The percentage with an OPCS varied from 70% of the caries free group to 92% of the group with caries, pathology and pain. However, there was no statistically significant association between state of the dentition and the presence of a PCS. Coincidence occurred in 56% of children. There was a statistically significant association between state of the dentition and coincidence (P=0·02). Children in pain were most likely to show coincidence (κ=0·81).