The separation of arsenic, lead, and bismuth from industrial antimony was investigated to determine suitable distillation conditions and product collection locations. Vacuum distillation was combined with fractional condensation, and the effects of distillation temperature and holding time on volatilization, product purity, direct yield, and impurity distribution were examined. Saturation vapor pressure, vapor collision frequency, and homogeneous nucleation calculations were used to interpret the separation behavior of antimony and its major impurities during evaporation and condensation. Industrial cathode antimony was treated in a vertical vacuum furnace equipped with a custom high-purity graphite crucible and staged graphite condensers. The feed contained 240.5, 410.0, and 30.5 mg/kg of As, Pb, and Bi, respectively. The evaporation zone and first condensation zone were separately temperature-controlled. Each experiment used a 200 g charge, a total furnace pressure of 100 Pa constrained by the available vacuum equipment, and a first-condensation-zone setpoint of 773 K, corrected using a temperature calibration ring. The temperatures at individual collection surfaces were not measured separately. The charge was heated to the target evaporation temperature over 2 h. Temperature experiments were conducted at 973, 993, 1 013, 1 033, 1 053, and 1 073 K for 6 h, whereas holding times of 4, 6, 8, 10, and 12 h were examined at 1 033 K. After cooling, the residue and deposits on each condenser wall and cover were collected and weighed separately. Impurity contents were measured by inductively coupled plasma mass spectrometry (ICP-MS). The deposit on the first condenser wall was designated as the target product, and its direct yield was calculated as the product mass divided by the initial feed mass. The results show that the calculated saturation vapor pressure of Sb increases from 65.26 Pa to 273.7 Pa between 973 K and 1 073 K, corresponding to a 4.19-fold increase. Pb and Bi show lower intrinsic volatility than Sb, whereas As shows a stronger tendency to enter the vapor phase. These differences support the retention of Pb and Bi at the evaporation end and the subsequent separation of As through staged condensation. Collision calculations use reference component partial pressures of 1, 10, and 20 Pa. At a fixed partial pressure, the calculated collision frequency decreases gradually with increasing temperature; at a fixed temperature, it increases proportionally with partial pressure. Cooling and increasing the Sb partial pressure both increase the calculated supersaturation. Component reference pressures used in the condensation calculations are distinguished from the total furnace pressure relevant to evaporation and vapor transport. Classical homogeneous nucleation calculations compare liquid and solid nuclei over 723–823 K. The liquid pathway represents supercooled liquid nuclei because this temperature range is below the bulk melting point of Sb. At an Sb partial pressure of 10 Pa, cooling from 823 K to 723 K reduces the liquid critical radius from 2.17 nm to 0.667 nm and the nucleation barrier from 647 kB T to 70.2 kB T, where kB denotes the Boltzmann constant and T denoted thermodynamic temperature. At 773 K and 10 Pa, the liquid and solid critical radii are 1.02 and 1.34 nm, and the corresponding barriers are 153 and 423 kBT. The liquid pathway has a lower barrier under the adopted interfacial-energy parameters. This comparison describes the relative tendencies of the two modeled pathways; it does not establish the actual phase of the initial nuclei. In the temperature series, volatilization increases from 4.64% at 973 K to 34.26% at 1 033 K and remains above 34% at 1 053–1 073 K. At 1 033 K, the first condenser-wall product reaches a purity of 99.9924% and a mass yield of 28.61%. Further heating reduces the target product yield to about 26% and its purity to approximately 99.962% at 1 073 K. The additional evaporation at higher temperatures thus does not translate into a corresponding increase in collected high-purity product. At 1 033 K, extending the holding time increases the volatilized fraction, which exceeds 90% after 12 h. Product direct yield increases overall and reaches 82.33% at 12 h, the highest value in the time series. Product purity varies with time and remains relatively high at both 6 and 12 h. A holding time of 6 h is selected for the regional composition analysis after considering product purity and processing duration. This condition provides high-purity antimony within a shorter holding period, whereas 12 h favores greater recovery per batch. At 1 033 K with a first-condensation-zone setpoint of 773 K and a holding time of 6 h, the first condenser-wall product reaches 99.9924% purity at a direct yield of 28.61%. Its As, Pb, and Bi contents are 36.91, 8.19, and 4.12 mg/kg, respectively. The Pb content of the residue is approximately 600 mg/kg, compared with 410.0 mg/kg in the feed. As contents reach approximately 5 800 mg/kg on the second condenser wall and 2 200 mg/kg on its cover. The first condenser cover also contains more As than the first condenser wall. These distributions demonstrate the need to collect deposits from different regions and surfaces separately. The combined experimental and calculated results support a separation mechanism involving selective evaporation followed by spatial separation during condensation. Preferential retention at the evaporation end restricts Pb and Bi transfer, while downstream enrichment of As reduces its incorporation into the target Sb deposit. Independent collection from the first condenser wall enables the preparation of antimony with a purity above 99.99%. Suitable evaporation intensity and separate collection of As-rich condensates are the principal operating requirements identified for this process. The limited degree of evaporation at 100 Pa contributes to the low 6 h product yield. Improving the vacuum system to lower the absolute pressure is expected to enhance evaporation and increase product yield when sufficient selective condensation is maintained.