Standard of care formulae for dose calculation for yttrium-90 hepatic radioembolization (RE) result in a wide range of efficacy and toxicity. A retrospective analysis was performed on patients treated with resin microspheres to calculate actual absorbed dose and to correlate this with clinical toxicity and response.All 247 patients treated with RE at a single center were retrospectively reviewed, and patients with hepatic metastases from colorectal carcinoma who received single session whole liver treatment with resin microspheres were specifically studied. Dose calculations were based on body surface area (BSA) and fractional liver involvement, according to the ‘BSA-method’. Liver volumes and whole liver absorbed doses were calculated and correlated with toxicity (NCI-CTCAE version 4.02) and radiographic response (WHO; RECIST).Forty-five patients were identified. Despite using the standard formula, the administered activity (range 0.85 - 2.58 GBq) did not correlate with the absorbed liver dose (median 49 Gy; range 30 - 75 Gy). Liver weight was highly variable (median 1741; range 941 - 3423 gm) and correlated strongly with the liver absorbed dose (p < 0.001). Patients who received liver absorbed doses > 50 Gy experienced more toxicity and adverse events (> grade 2 liver toxicity: 46% versus 17%; p < 0.05) but also responded better to the treatment than patients who received < 50 Gy (disease control 88% versus 24%; p < 0.01). Individual correlations were found between the liver absorbed dose and changes in bilirubin, liver enzymes and albumin, as well as changes in size of target lesions (p < 0.05).Dose-response and dose-toxicity relationships support using liver volume rather than BSA to prescribe the administered dose activity.