Since the efficiency of percutaneous ablation techniques in treating colorectal liver metastases is dependent on tumor size, the aim of this study was to verify the accuracy of computed tomography or magnetic resonance in estimating the maximum diameter of colorectal liver metastases by comparing these findings with those of pathology in a series of patients who underwent liver resection. Radiological and pathological tumor measurements in 39 patients operated for 69 colorectal liver metastasis were recorded. The radiological measurement was performed by magnetic resonance in 40 tumors (23 cases) and by computed tomography in 29 tumors (16 patients). The mean difference between pathological and radiological sizes was 0.4 cm (p < 0.001). Radiological size was smaller than pathological size in 60.9% of tumors, equal in 16% and bigger in 23% of tumors. The results indicated that radiology significantly underestimates the diameter of liver metastasis. To avoid on-site recurrence after percutaneous ablation therapies due to inaccurate radiological measurement, a radiological size of colorectal metastasis up to 2.5 cm should be considered as selection criteria for this treatment.