Gastro-oesophageal reflux disease (GERD) is difficult to diagnose without invasive testing. Peptest (RD Biomed, Hull, UK) is a recently marketed diagnostic tool which aims to quantify salivary pepsin as a marker of reflux, providing a rapid alternative to invasive procedures. To evaluate optimal timing for sampling, and to evaluate the accuracy of Peptest against an independent measure. Thirty diagnosed GERD patients (12 female, mean age 49 [range 20-72]) and 20 asymptomatic subjects (14 female, mean age 56 [range 21-56]) were subject to diurnal saliva sampling, with additional samples for 60 minutes following self-reported reflux symptoms and triggering of a proximal reflux alarm. Saliva samples were split and were analysed by both Peptest and ELISA with operators for each blinded to sample identity. Salivary pepsin was detectable in most patients and most volunteers. Peptest scores were significantly lower for patients than controls (P<0.005). ELISA scores showed no difference between patients and controls. There was no effect of diurnal sampling time (P=0.75) or time after symptoms (P=0.76) on Peptest readout. There was no correlation between Peptest and Pepsin ELISA (P=0.55); Bland-Altman analysis suggested no agreement between the tests (P=0.414). Receiver-operator curve suggests that neither Peptest (P=0.3328) nor pepsin (P=0.4476) is useful for predicting GERD. Salivary pepsin is not a reliable tool for the diagnosis of GERD.