Human can easily discriminate between randomly arranged and regularly arranged visual patterns. However, it is unclear what neural mechanisms involve such discrimination. Yamada et al. (2013) found that human observers can adapt to randomness of two-dimensional visual patterns; that is, perceived randomness decreased/increased after prolonged exposure of high/low physical randomness. The pattern randomness aftereffect was selective to the orientation of the pattern, but was not selective to the contrast polarity. Based on these findings, Yamada et al. hypothesized that the lateral occipital complex (LOC) is involved in the perception of pattern randomness. This study tested their hypothesis using fMRI. During the fMRI scanning, participants (n = 18) performed randomness judgment (RJ) and contrast judgment (CJ) for identical visual patterns. Behavioral results showed no difference between the tasks in the accuracy rates [0.91 ± 0.02 (mean ± SEM) for RJ, 0.91 ± 0.02 for CJ, P = 0.98] or reaction times (952.6 ± 45.7 ms for RJ, 930.8 ± 36.4 ms for CJ, P = 0.27). Brain imaging results showed that the RJ > CJ contrast revealed multiple activation regions, whereas the CJ > RJ contrast revealed no significant activation [extent threshold: P = 0.05 (FWE corrected), height threshold: P = 0.001 (uncorrected)]. The RJ-specific activity was observed in the right occipito-temporal region, extending from the fusiform gyrus to the inferior occipital and temporal lobes, which supports the LOC hypothesis. Similar activity was observed in the left hemisphere, although the activation cluster was smaller. In addition, the right premotor cortex and the bilateral inferior parietal lobes, extending to the primary somatosensory cortices (S1, BA 2/3) were also more activated during RJ. Thus, as predicted, the LOC was found to be a neural correlate for pattern randomness judgment. In addition, neural correlates were also observed in S1, along with the premotor and posterior parietal areas. The S1 activation may indicate that the brain applies the tactile area to processing for visual pattern randomness via LOC (cf., Amedi et al. 2001).