Abstract Purpose Positron emission tomography imaging of zirconium-89-labelled monoclonal antibodies (⁸⁹Zr-Immuno-PET) allows for visualisation and quantification of antibody uptake in tumours in vivo. Patlak linearization provides distribution volume (V_T) and nett influx rate (K_i) values, representing reversible and irreversible uptake, respectively. Standardised uptake value (SUV) and tumour-to-plasma/tumour-to-blood ratio (TPR/TBR) are often used, but their validity depends on the comparability of plasma kinetics and clearances. This study assesses the validity of SUV, TPR and TBR against Patlak K_i for quantifying irreversible ⁸⁹Zr-Immuno-PET uptake in tumours. Methods Ten patients received 37 MBq 10 mg ⁸⁹Zr-anti-EGFR with 500 mg/m² unlabelled mAbs. Five patients received two doses of 37 MBq ⁸⁹Zr-anti-HER3: 8–24 mg for the first administration and 24 mg–30 mg/kg for the second. Seven tumours from four patients showed ⁸⁹Zr-anti-EGFR uptake, and 18 tumours from five patients showed ⁸⁹Zr-anti-HER3 uptake. SUV_peak, TPR_peak and TBR_peak values were obtained from one to six days p.i. Patlak linearization was applied to tumour time activity curves and plasma samples to obtain K_i. Results For ⁸⁹Zr-anti-EGFR, there was a small variability along the linear regression line between SUV (− 0.51–0.57), TPR (− 0.06‒0.11) and TBR (− 0.13‒0.16) on day 6 versus K_i. Similar doses of ⁸⁹Zr-anti-HER3 showed similar variability for SUV (− 1.3‒1.0), TPR (− 1.1‒0.53) and TBR (− 1.5‒0.72) on day 5 versus K_i. However, for the second administration of ⁸⁹Zr-anti-HER3 with a large variability in administered mass doses, SUV showed a larger variability (− 1.4‒2.3) along the regression line with K_i, which improved when using TPR (− 0.38–0.32) or TBR (− 0.56‒0.46). Conclusion SUV, TPR and TBR at late time points were valid for quantifying irreversible lesional ⁸⁹Zr-Immuno-PET uptake when constant mass doses were administered. However, for variable mass doses, only TPR and TBR provided reliable values for irreversible uptake, but not SUV, because SUV does not take patient and mass dose-specific plasma clearance into account.