In helping consumers to make a wise decision in purchasing a personal computer (PC), the use of multi-criteria decision analysis methods is a way to provide ranking of the attributes of PCs and to construct the computer preference index (CPI). This paper employs the Rank Ordered Centroid (ROC) method to do the ranking. The findings reveal that the most important attribute is the CPU, followed by the hard drive, the price, the memory card, the warranty, the size, the screen resolution, the Ethernet, the weight and the DVD. Whilst, the CPI is constructed by using a Multi-Criteria Decision Making (MCDM), called the Simple Additive Weighting (SAW) method. Analysis of data from 25 PCs from four brands, Toshiba, Dell, HP Compaq and Acer, presents the CPI. The resulted CPI shows that two HP Compaq models, Presario V3632 and V3653 are preferred most, while, Acer Aspire 4920-5AA0516MI is the least preferred.