The pH of Tarn Head Beck, an acidic upland stream in the English Lake District, was continuously recorded for a period of 6 months. Preliminary measurements showed that electrodes should be completely immersed to avoid problems associated with the difference in temperature between water and air. For a site without mains power good results were obtained by deploying electrodes within the stream in purpose-built units. There was good agreement between these directly deployed electrodes and electrodes immersed in an auxiliary flow in an instrument box on the bank of the stream. Spot measurements taken from the continuous record also compared well with dip samples collected into bottles and subsequently measured in the laboratory. There were no problems associated with long-term drift under field conditions, even though preliminary laboratory experiments had indicated that the stability of the electrodes would be inadequate. Electrodes appear to perform more reproducibly after long-term immersion in a relatively constant medium. Measurements of hydrogen ion fluxes over 14 day periods obtained using an automatic integrating sampler and the total flow were compared with measurements calculated from continuously recorded pH and flow. Although agreement was good at low and medium flow, the integrating sampler results appeared to underestimate results at high flow. For annual budget studies, the use of event triggered samplers is recommended.