A group of 72 children (mean age: 21.7 months, range: 14 days-19 years) with symptoms of gastro-oesophageal reflux were investigated by 22 h pH monitoring. Using a Proxeda software, we compared, in the same patient, the specificity and sensitivity of pH monitoring during 3 h, 6 h, 12 h, 12 nocturnal hours and 3 postprandial hours, as well as the influence of position and the sleep and alert periods. Results showed that all the short pH monitorings were statistically less sensitive than 22 h pH monitoring (P < 0.025). As regards specificity, only the 12 nocturnal hours pH monitoring was not statistically different from the 22 h pH monitoring. Gastro-oesophageal reflux was more frequent when the patient was awake than during sleep. pH monitoring seemed more reliable in the recumbent than in the upright position. We conclude that long-term pH monitoring (22 h) is the test of choice to diagnose gastro-oesophageal reflux because it included sleep and alert periods as well as different positions.