Ketone ester (KE) drinks are sold commercially in the USA and elsewhere (the FDA has accepted a GRAS (generally regarded as safe) notification on the KEs we propose to study). KE drinks can have beneficial effects on metabolism. Unfortunately, KE drinks have a strongly unpleasant flavor which can trigger rejection. The current study will generate data to help inform KE drink formulations with improved flavor. First stage, trained subjects will rate the intensity of various sensation qualities (e.g., bitter, sour, metallic, and astringent) from KE drinks to help understand the nature of the unpleasant flavor and the particular sensory systems involved (e.g., taste vs. smell). Then, subjects will taste KE drinks with and without various flavor modifiers. Subjects will be healthy adults (aged 21-45) without known food allergies.

Inclusion Criteria:

1. Good general health (self report)
2. Men and women
3. Age from 21 to 45

Exclusion Criteria:

1. Current (or within the last four weeks) acute illness, such as cold or flu
2. Currently active allergies
3. Regular use of medication (over the counter or prescription), except for birth control.
4. Any history of food allergy or sensitivity (particularly to KE drinks or MSG). Any history of sensitivity to 6-Propylthiouracil (bitter compound frequently used to assess individual differences in sensitivity to bitterness).
5. Pregnant, nursing, or trying to become pregnant during the study (pregnancy can affect flavor perception)
6. Chronic illness (any ongoing condition that is serious or required medical monitoring), including kidney problems, liver problems, cardiovascular problems (e.g., heart disease or high blood pressure), diabetes, HIV infection, or other ongoing illness). Diabetes or other metabolic issues could be important for safety as well as quality of data.
7. History of taste problems (absent or abnormal sense of taste)
8. Weigh less than 100 pounds (45.4 kilograms). This is to help ensure that the amount of KE drink consumed remains below the single serving in the GRAS (generally regarded as safe) notice which the FDA has accepted.

If individual subject data are shared, the data will be de-identified first (no identifying information will be shared with other researchers).

Samples presented will not be identified (there will be no codes or visual cues regarding whether a model drink does or does not contain a flavor modifier). However, true blinding is impossible since flavor will differ between samples.

Investigators will not be blind to sample contents. However, responses will be collected and recorded using an automated computer program, which offers some protection against possible experimenter bias.