Forty bottle-fed babies, 4-12 weeks old, with clinical gastro-oesophageal reflux were studied. Continuous 24 h oesophageal pH monitoring in a prone position demonstrated a gastro-oesophageal reflux with all of the following parameters: reflux index, duration of the longest reflux episode, number of reflux episodes in 24 h, number of reflux episodes greater than 5 min in 24 h. Positional therapy (prone-antitrendelenburg position), applied to all infants, resulted in a normalization of these parameters in ten of them. The remaining 30 infants were treated with milk-thickening agents, as recommended by Carre. Nearly all (N = 25) showed an important clinical improvement. A third pH monitoring was performed after 10-14 days of treatment. In six infants the results were completely within normal ranges. In 24 infants a decrease in the number of reflux episodes was demonstrated, with a comparable reflux index and number of long lasting reflux episodes. The duration of the longest reflux episode however increased significantly (P less than 0.001). Drugs (domperidone, Gaviscon) added to the milk-thickening agents in these 24 children, led to normalization of pH tracings. Clinical symptoms were less severe or disappeared in all infants but one. We conclude that if positional therapy (prone-antitrendelenburg) does not correct gastro-oesophageal reflux in infants, pharmacological treatment should be applied. Milk-thickening agents alone can be effective in individual cases but should be prescribed with care as they can lead to more occult gastro-oesophageal reflux with episodes of longer duration, increasing the risk of oesophagitis or respiratory distress.