Diffusion-weighted imaging (DWI) in human transplantation was regarded as a promising indicator for dectecting graft dysfunction. By applying the intravoxel incoherent motion (IVIM) theory, the bi-exponential model fitting with a number of b-values yielded the standard apparent diffusion coefficient (ADC_st), the “perfusion fraction” (f ), and the “slow diffusion” (ADC_slow)(1). Besides perfusion of microcirculation, other biophysical mechanisms may also play a substantial role in contributing to ADC changes, especially membrane permeability to water which is mainly determined by aquaporins (AQP)(2). T. Tourdias et al reported that ADC might also indirectly reflect quantitative water channel expression(3). So in this study, we performed DWI with multi-b values in renal allografts to determine whether IVIM parameters and AQP ADC values vary in different estimated glomerular filtration rate (eGFR) levels and correlate with clinical indicators.Respiratory-triggered DWI was performed on a 3.0 Tesla MR scanner in 12 renal allograft recipients. Because of the small sample, we divided the patients into 2 groups according to allograft function: Group 1, eGFR≥60 ml/min/1.73m² (n =6; 1 women, 5 men; mean age, 36.2±7.3 years); Group 2, eGFR<60 ml/min/1.73m² (n =6; 2 women, 4 men; mean age, 38.5±11.3 years). Multi-b DWI (b=0, 10, 30, 50, 70, 100, 150, 200, 400, 800, 1000, 1500, 2000, 2500, 3000, 3500, 4000) was performed in all renal allograft recipients and IVIM parameters (ADC_st, ADC_slow and f ) and AQP ADC values were calculated. The DWI images and parameter maps are shown in Fig 1. (A-C: a patient with eGFR 88.71 ml/min/1.73m²; D-F: a patient with eGFR 15.22 ml/min/1.73m²; A, D: DW image obtained with a b value of 100 s/mm²; B, E: ADCst map; C, F: AQP ADC map). ALL IVIM parameters (corticomedullary ADC_st, ADC_slow and f ) and cortical AQP ADC values differed significantly between different eGFR levels (p<0.05). Cortical ADC_st (r=0.789, p=0.002), cortical f (r=0.773, p=0.003), medullary ADC_st (r=0.828, p=0.001), medullary ADC_slow (r=0.621, p=0.031), and medullary f (r=0.777, p=0.003) showed significantly positive correlations with eGFR, while AQP ADC (r= -0.841, p=0.001) showed a negative correlation with eGFR(Fig 2).In our study, all parameters except medullary AQP ADC could detect renal allografts with different eGFR levels. According to the IVIM theory, ADC integrates the effects of both diffusion of water molecules and microcirculation of blood in capillaries. By separating diffusion and perfusion, we could observe each component’s contribution to the changes of ADC values. Furthermore, the exchanges of intracellular water molecules with extracellular’s may have impact on ADC. Our study showed strong correlation between cortical AQP ADC and eGFR, indicating that AQP ADC maybe a potential tool for reflecting the expression of aquaporins. In conclusion, the present investigation demonstrates that IVIM DWI and AQP MR imaging are promising new techniques for functional evaluation of renal allografts.