Exhaustive mode decision has been exploited in multiview video coding for effectively improving the coding efficiency, but at the expense of yielding much higher computational complexity. In this paper, a fast mode decision algorithm, called the mode correlation-based mode decision (MCMD), is proposed to speed up the encoding process by reducing the number of the modes required to be checked. In our approach, all the prediction modes are first categorized into five motion-activity classes, and only one of them will be chosen to identify the optimal mode in a hierarchical manner, as follows. For each macroblock (MB), the proposed MCMD algorithm always begins with checking whether the rate-distortion cost computed at the SKIP mode (i.e., Class 1) is below an adaptive threshold for providing a possible early termination chance. If this early termination condition is not met, one of the remaining four motion-activity classes will be chosen for further mode checking according to the analysis of the predicted motion vector (PMV) of the current MB. The above-mentioned adaptive threshold and PMV are derived by exploiting the mode correlation between the current MB and a set of adjacent MBs (i.e., region of support) in the current view and its neighboring view. Experimental results have shown that compared with exhaustive mode decision, which is a default approach set in the joint multiview video model (JMVM) reference software, the proposed MCMD algorithm achieves a reduction of the computational complexity by 73.39% on average, while incurring only 0.07 dB loss in peak signal-to-noise ratio (PSNR) and 2.22% increment on the total bit rate.