This study was undertaken to evaluate whether intrigastric pH measurements and bile salt concentrations show synchronous fluctuations, in both postprandial and fasting states. It was performed on 5 normal volunteers and 17 non-operated patients suspected of having duodenogastric reflux. In controls the two variables showed different behaviour, especially after eating, while in suspected refluxers they presented more synchronous elevations both in postprandial states and in the nocturnal period. The first derivatives of both pH and bile salt interpolations using Fourier's truncated Series give evidence that these two variables behave differently in the time domain of the two groups. In fact, their first derivatives show equal signs for 521 min (36.2%) out of the 1440 of the study in controls (p less than 0.05) and for 1024/1440 min. (71.1%) in suspected refluxers (p less than 0.005). The difference between the time elapsed in min with equal sign derivatives in the two groups is significant (p less than 0.005). Despite this more coherent behaviour of pH and bile salt fluctuations in patients with suspected reflux, the two variables do not present coincident numerical values even in this population and therefore a close quantitative relationship between them can be excluded in both groups. As a consequence of these results, the use of bile salt concentrations seems inadequate to validate the fact that alkalinizations observed in pH-metry studies are related to episodes of duodenograstric reflux.