There currently is no clinically accepted noninvasive technique for detecting moderate blood loss. Clinicians instead normally rely on lagging indicators such as blood pressure and tachycardia. We propose to use changes in the morphology of the respiratory induced variation in the photoplethysmogram (PPG) to detect moderate hypovolemia in non-ventilated subjects. These changes were characterized by two statistically robust metrics that were developed to characterize the top and bottom envelope of the PPG. The first metric detects when the height of the top envelope becomes greater than the difference between the minimum of the top envelope and the maximum of the bottom envelope. The second metric robustly detects when the upper and lower envelopes synchronously rise or fall. The use of these metrics was then validated in nonintubated healthy volunteers with a Lower-Body Negative Pressure (LBNP) chamber which induces central hypovolemia by sequestering blood in the hips and lower extremities. Hypovolemia corresponding to sequestration of more than 1 liter of blood (LBNP > 60 mmHg) was consistently detected using these metrics before significant change in blood pressure, or tachycardia are observed.