Blood hemograms are an essential part of the clinical pathology diagnosis. This is especially important for chronic or acute conditions that require blood transfusions (e.g., anemia, cancer, hemophilia, and kidney and liver diseases). The capacity to monitor hemogram parameters using a wearable device without blood withdrawal is evaluated in this study using ultraviolet, visible, and near-infrared spectroscopy transmittance measurements simulating the wrist vein. Tylose gel and catheter tubes representing internal and near-surface wrist veins were studied to simulate different anatomical characteristics (e.g., skin tone, thickness, and depth of the monitored vein). The results showed that diffusive transmittance through the anatomical surrogate can correctly quantify the hemogram in terms of red blood cells (RBC), hemoglobin (Hgb), and discriminate levels of total white blood cells (WBC), achieving the following correlation (R) and total error (TE) against laboratory automated hematology machines: i. RBC: R=0.85 and TE=5.9%; ii. Hgb: R=0.81 and TE=5.25%; and iii. WBC: R=0.78 and TE=12.0%. These results demonstrate the possibility of performing a hemogram analysis without blood withdrawal using diffuse reflectance spectroscopy and artificial intelligence. Further research is necessary to evaluate the efficiency of different optical designs to be incorporated into a wearable bracelet system for noninvasive blood hemogram analysis.