Objective: The study was designed to determine the distance between the visible "ice-ball" and the lethal temperature isotherm for normal renal tissue during cryoablation. Methods: The Animal Care Committee approved the study. Nine adult swine were used: three to determine the optimum tissue stain and six to test the hypotheses. They were anesthetized and the left renal artery was catheterized under fluoroscopy. Under MR guidance, the kidney was ablated and (at end of a complete ablation) the nonfrozen renal tissue (surrounding the "ice-ball") was stained via renal artery catheter. Kidneys were explanted and sent for slide preparation and examination. From each slide, we measured the maximum, minimum, and an in-between distance from the stained to the lethal tissue boundaries (margin). We examined each slide for evidence of "heat pump" effect. Results: A total of 126 measurements of the margin (visible "ice-ball"-lethal margin) were made. These measurements were obtained from 29 slides prepared from the 6 test animals. Mean width was 0.75 ± 0.44 mm (maximum 1.15 ± 0.51 mm). It was found to increase adjacent to large blood vessels. No "heat pump" effect was noted within the lethal zone. Data are limited to normal swine renal tissue. Conclusions: Considering the effects of the "heat pump" phenomenon for normal renal tissue, the margin was measured to be 1.15 ± 0.51 mm. To approximate the efficacy of the "gold standard" (partial nephrectomy, ~98 %), a minimum margin of 3 mm is recommended (3 × SD). Given these assumptions and extrapolating for renal cancer, which reportedly is more cryoresistant with a lethal temperature of -40 C, the recommended margin is 6 mm. © 2012 Springer Science+Business Media, LLC and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE).