Electroretinography is commonly used to assess the functional integrity of the retina. There are many external variables that can influence the electroretinographic waveforms recorded, and it is important to be aware of these so as not to misinterpret their effects as abnormalities in retinal function. In this study we examined the effect of three different recording electrodes on the ERGs recorded from normal dogs. A bipolar Burian-Allen lens, a monopolar Dawson Trick Litzkow (DTL) fiber electrode, and a monopolar ERG-Jet lens electrode were compared. The effect of altering the distance of the reference electrode from the eye was also examined; using the ERG-Jet lens electrode, the ERG was recorded with the reference electrode placed over the zygomatic arch at 1, 3 and 5 cm caudal to the lateral canthus. The ERGs recorded with the bipolar Burian-Allen lens had significantly lower amplitudes, higher a-wave thresholds and a shallower initial a-wave slope, than those recorded by the two monopolar electrodes. Positioning the reference electrode further from the eye resulted in significantly higher amplitudes. Naka-Rushton fitting and calculation of retinal sensitivity (K) gave significantly different results between the Burian-Allen lens and ERG-Jet lens electrode with the reference electrode 5 cm from the lateral canthus. These results demonstrate that recording electrode type and distance of the reference electrode from the eye significantly affect the ERG tracings of the dog, and may alter the assessment of retinal function that can therefore be derived. Results obtained using these three different types of electrodes cannot be directly compared.