Percentage hepatic replacement (PHR) may be an important factor in assessing prognosis, staging and clinical trial evaluation. Most imaging techniques used to assess PHR rely on subjective interpretation. We have employed the technique of planimetry for the quantitative measurement of PHR from CT scans of patients with colorectal liver metastases. This micro-computer based system was assessed for precision by repeated estimates of CT images of livers with variable metastatic features. Precision was correlated with single and large metastases. Accuracy of CT-planimetry was assessed against cadaveric livers containing metastases for which PHR had been determined by water displacement. For the PHR range 1.2-53.5, the range of percentage error measured was between -12.8 and +8.51. Subjective assessment of PHR scans by individual surgeons and radiologists was varied and idiosyncratic. No significant difference between the two groups could be determined. The greatest errors in the subjective estimates of PHR occurred in the range 5-15%. It is concluded that quantitative methodology must be employed to estimate the percentage hepatic replacement.