OBJECTIVE The primary aim was to assess whether there were differences in symptoms, laboratory data, and oesophageal pH-metry between infants with primary gastro-oesophageal reflux and those with reflux secondary to cows' milk protein allergy (CMPA). METHODS 96 infants (mean(SD) age 7.8(2.0) months) with either primary gastro-oesophageal reflux, reflux with CMPA, CMPA only, or none of these (controls) were studied. Symptoms, immunochemical data, and oesophageal pH were compared between the four groups and the effect of a cows' milk protein-free diet on the severity of symptoms was also assessed. RESULTS 14 out of 47(30%) infants with gastro-oesophageal reflux had CMPA. These infants had similar symptoms to those with primary gastro-oesophageal reflux but higher concentrations of total IgE and circulating eosinophils (p < 0.005) and IgG anti-beta lactoglobulin (p < 0.003). A progressive constant reduction in oesophageal pH at the end of a feed, which continued up to the next feed, was seen in 12 out of 14 patients with gastro-oesophageal reflux secondary to CMPA and in 24 of 25 infants with CMPA only. No infants with primary gastro-oesophageal reflux and none of the controls had this pattern. A cows' milk protein-free diet was associated with a significant improvement in symptoms only in infants with gastro-oesophageal reflux with CMPA. CONCLUSIONS A characteristic oesophageal pH pattern is useful in distinguishing infants with gastro-oesophageal reflux associated with CMPA.