To evaluate the effect of pacifier use (nonnutritive sucking) on gastroesophageal reflux in infants, 48 infants younger than 6 months of age with pathologic reflux were prospectively evaluated with pH probe. In each infant, parameters of reflux were blindly quantified during paired periods in a cross-over design when pacifier use was either encouraged or prohibited. To determine whether positioning was a factor in the effect of nonnutritive sucking on reflux, 24 of the infants were studied seated and 24 studied prone. Pacifier use significantly affected only the frequency of reflux episodes, increasing it in prone infants from 7.2 +/- 1.1 to 12.8 +/- 2.3 episodes/120 min postprandially (p = 0.040) and decreasing it in seated infants, from 21.1 +/- 3.1 to 14.8 +/- 2.6 postprandially (p = 0.003) and from 17.3 +/- 4.8 to 5.9 +/- 0.9 in the fasting period (p = 0.035). It did not significantly affect the clearance of reflux episodes or the total reflux time. These results suggest that infants with pathologic reflux frequency might best avoid pacifier use while in the beneficial prone position. When seated position is necessary, the pacifying effects of nonnutritive sucking may be useful in decreasing reflux events as well as in reducing crying behavior.