Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) are non-contrast MRI measurements of water mobility that reflect tissue microenvironment. These measurements have gained research interest for the assessment of breast cancer characteristics ^1,2. As we continue to advance imaging techniques for assessing tumors and their response to therapy, it is important to understand the underlying variability of these imaging measurements (both within- and between-subjects) in order to correctly interpret the importance of changes in patients and determine the optimal course of treatment. The purpose of this study was to assess breast tissue measurements using diffusion MRI techniques in repeated studies to evaluate the variability of ADC and FA measurements within- and between-subjects.With institutional review board approval, an estimation of variability and repeatability of ADC and FA measurements in normal breast tissue was conducted in 13 normal pre-menopausal subjects. MR diffusion imaging was performed on a 1.5T scanner (Signa LX, GE Healthcare, WI) using a bilateral 8-channel phased array breast coil (Invivo, FL), utilizing an axial, fat suppressed, echo-plane imaging (EPI) sequence. ADC was measured with a 3 direction, 4 b-value (0, 100, 600, 800 s/mm2) DWI acquisition as defined in the ACRIN 6698 study protocol ³ and along with FA, it was also measured with a 6 direction, 2 b-value (0, 600 s/mm²) DTI acquisition. In the test/retest protocol, all subjects were removed from the scanner and repositioned between the test and retest data acquisitions. The same day test-retest design was used to minimize hormonal induced variation in the breast tissue measurements ⁴. Parametric maps were generated for ADC and FA based on methods previously described ^5,6. Between- and within-subject variability of measurements was assessed by the coefficient of variation (CV = standard deviation, SD, divided by the mean). Scatter-plots and Spearman’s rank correlation were used to assess the agreement between test and retest results. Bland-Altman plots of difference vs. mean of measurements from test-retest were used to assess the repeatability of measurements. The coefficient of repeatability (CR =1.96*SD_diff, where SD_diff = standard deviation of measurement difference from test and retest) was used to assess the reliability of the measurements of normal breast tissue across subjects ⁷.The between-subject variation (bCV) of DWI derived ADC ranged from 14.5-14.7%, DTI derived ADC was 12.2-15.3% and FA was 26.7-27.2% at test and retest. The within-subject CV (wCV) was 3.83% for ADC from DWI, 4.88% from DTI and 11.2% for FA. A strong correlation between ADC test and retest measurements was observed in both DWI and DTI acquisitions with correlation coefficient of 0.965 (95% CI (0.80, 1.00), p< 0.0001) and 0.940 (95% CI (0.75, 1.00), p< 0.0001) respectively. A weaker correlation was observed in FA (ρ= 0.747, 95% CI (0.24, 0.98), p = 0.05). The CR of ADC from DWI was 0.219x10^-3 mm²/s (11%; average ADC_test = 1.99 x10^-3 mm²/s and ADC_retest = 1.98 x10^-3 mm²/s) and was 0.295 x10^-3 mm²/s from DTI (14%; average ADC_test = 2.09 x10^-3 mm²/s and ADC_retest = 2.10 x10^-3 mm²/s). The CR of FA was 0.051 (32%) with average FA_test = 0.159 and FA_retest = 0.168. Scatter-plots and Bland-Altman plots of ADC and FA measurements from test and retest are shown in Figure 1 and 2 respectively. The plots further indicate the improved correlation and repeatability for both Mean ADC measures vs. Mean FA.Both repeatability and within-subject variance of DTI derived ADC and FA were in agreement with previously published findings ⁸. Compared to DWI, DTI acquired ADC had higher wCV and CR values. Influence of tissue location and b-value on breast DTI measurements may account for the increased variance of ADC values ⁸. The weaker test-retest correlation and high CR found in FA indicates that the FA measurement is both less repeatable and less reliable compared with ADC and changes observed in patients using FA as an imaging metric should be interpret with caution.