Diffusion kurtosis imaging (DKI), evaluating the non-Gaussianity of water diffusion, has been demonstrated to be sensitive biomarker in many neurological diseases. Researchers have been trying to optimize the diffusion schemes and protocols for accurate measurements of DKI parameters. However, number of repetition is one of the factors, but people is trying less to investigate it. In this study, normal rats were performed using two different diffusion scheme protocols and with different repetitions. Both mean diffusivity (MD) and mean kurtosis (MK) were quantified for further comparison. Animals: Three naïve adult Sprague-Dawley rats (250 to 300 g) were used for evaluation of mean diffusivity (MD) and mean kurtosis (MK) using two protocols with different number of b-values and diffusion-encoding scheme for different repetitions. MRI measurements: Diffusion weighted imaging was performed on a Bruker 7T/30cm Biospec scanner utilizing a single-shot, spin-echo, echo-planar imaging sequence. One protocol was applied along with conventional 6 diffusion-encoding directions with total 15 b-values: 250, 400, 500, 600, 700, 800, 1000, 1100, 1250, 1350, 1500, 1750, 2000, 2250, 2500 s/mm2. Another protocol was applied along with 30 diffusion-encoding directions with 3 b-values: 0, 1000, and 2000 s/mm2. All images were obtained with other sequence parameters of TR 2.5 sec, TE 40 ms, D 16 ms, d 4 ms, Average 1, NEX 8, slice thickness 2 mm, field-of-view 2.5 cm ´ 2.5 cm, data matrix 96 ´ 96 (without zero-filled). Five b=0 images were also employed. Each animal was underwent DKI examination using both protocols at the same time for comparison. Acquisition time per repetition per protocol was around 8 minutes. Data was analyzed with DKI model analysis packages developed in-house with Matlab. The MD and MK were obtained by fitting all diffusion weighted images to DKI model, ln[S(b0)] = ln[S(0)]-bD+(1/6)b2D2K (Jensen et al., 2005). The fractional anisotropy (FA) could be also calculated. ROI analysis: motor cortex and striatum were respectively selected in the anterior part of the normal rat brain based on individual FA map (shown in Figure 1). ROI analysis was performed on MD and MK maps derived from diffusion weighted images by both protocols using total 1, 2, 4 and 8 repetitions. Statistical analysis: data are presented as mean ± standard deviation. Two sample t-test was performed with Excel 2010. Figure 2 shows the representative MD and MK maps of a normal rat brain using different diffusion schemes (15 b-values with six diffusion directions vs. 3 b-values with thirty diffusion directions) with 1, 2, 4, and 8 repetitions. MD map provides different contrast sensitivity of brain microstructures, as compared with MK map in the same image section. MD maps show similar not only between two protocols but also different repetitions. MK maps show darker and slightly different tissue contrast using the 15 b-values protocol, as compared with the 3 b-values protocol. Also, we observe some possible failure calculation of voxels such as in corpus callosum regions, shown in MK maps in our case. Figure 3 show the quantitative results of MD and MK values using different protocols and repetitions for comparison. MD values of either control cortex or striatum was ~0.75 μm2/ms which are calculated using different protocol and repetitions. MK values of both regions, however, shows higher by using 3 b-value protocol, compared to 15 b-value protocol (MK of cortex vs. striatum: 0.55 vs. 0.54 by 15 b-value protocol, and 0.76 vs. 0.79 by 3 b-values protocol).No significant difference was found in MD and MK of both regions from same scheme with different repetitions in Figure 3, suggesting the protocol with one repetition provides good image quality for DKI analysis in this case. MD of both regions show no difference between two protocols as well. It may due to similar tissue microstructures, gray matter, in selected cortex and striatum regions. In contrast, MK were significantly smaller and with relatively larger standard deviation using 15 b-value than 3 b-value protocols in both regions, indicating a possible underestimation of MK. Although our results indicated that no difference found in MD and MK by repetitions in this image design, it is still worth a try to evaluate image data with higher in-plan resolution and thinner thickness, usually used for disease of rodent study for accurate estimation. In addition, the protocol with 3 b-value and thirty direction may be a better one for MK measurements.