Many general contractors obtain a majority of projects based on the low-bid award system. A major objective of a competitive bidding model is to determine an optimum markup size so as to maximize the contractor's long term profit. The new bidding model with explicit consideration of correlation is proposed since this imporant parameter is not considered in existing bidding models. In this study, the existence of a positive correlation coefficient between any two competitors' bid ratios was demonstrated. After that, a new competitive bidding model was proposed, and a statitstical method in a Bayesian framework was developed. The significance of correlation on probability of winning and optimum markup decisions was investigated. For an illustration purpose, a case study of a bidding Data Set from actual projects was conducted. It has been found that as correlation increases, the probability of winning will increase, and hence an increased optimum markup can be used. In comparison with Friedman and Gates models, the proposed model with consideration of correlation coefficient derives different value of optimum markup which is closer to the real situation of the construction market since the correlation coefficient among competitors' bid ratios is considered.