Electrogustometry is well established as a clinical tool for the estimation of taste detection thresholds. Nevertheless, the user is sometimes unaware of the impact of superficially minor procedural and psychophysical factors upon the reliability and comparability of threshold estimates. The inherent strengths and limitations of the procedure are outlined, and aspects of the control and specification of the stimulus that moderate threshold measures are discussed. In addition, threshold estimates from two individuals with severe unilateral taste loss are used to illustrate the level at which anodal dc current may elicit common, rather than taste, sensation. Where chorda tympani section is complete and historical (older than 7-14 days), very high stimulus levels, conservatively over 5 micro A/mm2 (100 micro A linear current with a 5-mm diameter electrode), are required to activate trigeminal responses.