Objectives We investigated the feasibility and reproducibility of free-breathing motion-corrected multiple inversion time (multi-TI) pulsed renal arterial spin labelling (PASL), with general kinetic model parametric mapping, to simultaneously quantify renal perfusion (RBF), bolus arrival time (BAT) and tissue T 1 . Methods In a study approved by the Health Research Authority, 12 healthy volunteers (mean age, 27.6 ± 18.5 years; 5 male) gave informed consent for renal imaging at 3 T using multi-TI ASL and conventional single-TI ASL. Glyceryl trinitrate (GTN) was used as a vasodilator challenge in six subjects. Flow-sensitive alternating inversion recovery (FAIR) preparation was used with background suppression and 3D-GRASE (gradient and spin echo) read-out, and images were motion-corrected. Parametric maps of RBF, BAT and T 1 were derived for both kidneys. Agreement was assessed using Pearson correlation and Bland-Altman plots. Results Inter-study correlation of whole-kidney RBF was good for both single-TI (r 2 = 0.90), and multi-TI ASL (r 2 = 0.92). Single-TI ASL gave a higher estimate of whole-kidney RBF compared to multi-TI ASL (mean bias, 29.3 ml/min/100 g; p <0.001). Using multi-TI ASL, the median T 1 of renal cortex was shorter than that of medulla (799.6 ms vs 807.1 ms, p = 0.01), and mean whole-kidney BAT was 269.7 ± 56.5 ms. GTN had an effect on systolic blood pressure (p < 0.05) but the change in RBF was not significant. Conclusions Free-breathing multi-TI renal ASL is feasible and reproducible at 3 T, providing simultaneous measurement of renal perfusion, haemodynamic parameters and tissue characteristics at baseline and during pharmacological challenge. Key points • Multiple inversion time arterial spin labelling (ASL) of the kidneys is feasible and reproducible at 3 T.• This approach allows simultaneous mapping of renal perfusion, bolus arrival time and tissue T 1 during free breathing.• This technique enables repeated measures of renal haemodynamic characteristics during pharmacological challenge.