This study proposed to demonstrate a classically conditioned blood sugar decrease in humans and to clarify its relevance for binge eating. Six conditioning trials were run in healthy females. The conditioned stimulus (CS) was a compound peppermint flavor/fragrance, whereas the unconditioned stimulus (UCS) consisted of 50 g of oral glucose. Control subjects received an aspartame drink as the UCS. Ad lib glucose intake, blood parameters, and subjective craving were monitored before and after conditioning. Results showed that the experimental group failed to show conditioned blood sugar and glucagon decreases or C-Peptide increases. Although an increased insulin response was found in the experimental group, the effect size did not exceed that of spontaneous fluctuations. No increases in craving for sweet substances were found. An impressive increase (mean: 78%) in glucose intake after conditioning was found in both conditions, as well as in a subsequently run third condition with plain water as the UCS. The increased glucose intake probably resulted from an initial neophobia to the laboratory setting that subsided as subjects had experienced more lab sessions. Importantly, because no conditioned hypoglycemia occurred in the present study, its relationship with subjectively experienced craving for sweet substance could not be determined.