Background: To evaluate the association between various radiologic-pathologic findings and oncogenic Kirsten-ras (KRAS) mutation in patients with primary rectal cancer. Methods: Seventy-five patients with primary rectal cancer who had undergone rectal magnetic resonance imaging (MRI) were included. The rectal MRI consisted of T2-weighted images in three planes, pre-and postcontrast-enhanced T1-weighted images, and axial diffusion-weighted images (b factors, 0, 1,000 s/mm 2). Two radiologists reviewed the MRI scans and measured the axial and longitudinal tumor lengths (LTLs), apparent diffusion coefficient (ADC), and relative contrast enhancement [signal intensity (SI) difference of tumor on pre-and post-contrast T1WI/SI of tumor on pre-contrast T1WI]. The associations among the qualitative data (tumor stage, node stage, lymphatic invasion, venous invasion, and perineural invasion), quantitative data (tumor length, ADC, relative contrast enhancement) and KRAS mutations were statistically analyzed by Fisher's exact test for the qualitative data and by the Mann-Whitney U test for the quantitative data. An area under receiver operating characteristic curve (AUC) was considered as the diagnostic performance for the prediction of KRAS mutation. Molecular-biologic results served as the reference standard. Results: The ratio of axial to LTL in the KRAS-mutant group (n=41) was higher than that in the wild-type group (n=34) (0.29±0.15; 0.22±0.08, P=0.0117). The AUC was 0.640 (95% CI, 0.520 to 0.747, P=0.0292) with an estimated maximum accuracy of 64%. The mean ADC of the mutant group was not significantly different from that of the wild-type group [(0.95±0.17)×10 −3 mm 2 /s; (0.96±0.17)×10 −3 mm 2 /s, P=0.6505]. The relative contrast enhancement showed no significant difference between the two groups (1.66±0.93, 1.35±0.84, P=0.1581). The other qualitative findings also did not show any significant difference (P>0.05). Conclusions: The ratio of axial to LTL showed a significant difference according to KRAS mutation in patients with primary rectal cancer. However, it showed a low accuracy of 64% for prediction of KRAS mutation.