The main premise of transaction cost economics is that project cost is not only production costs but there are also additional costs resulting from transactions between parties. Transaction costs include the costs of preparing the bidding package, estimating and drawing up a contract, administering the contract, dealing with any deviations from contract conditions and any contractual problems include claims, change orders, and disputes. The main objectives of this paper are to establish the factors that affecting transaction costs in construction projects in Egypt, assessing its value in Egyptian construction projects, and to provide a mathematical model that can predict transaction costs value accurately. The study was conducted on various types of construction projects in Egypt, which formed the sample size. To collect data, a questionnaire was used that was personally administered to the respondents, using emails and conducting semi-structured interviews. The collected data were analyzed by using SPSS program, and a software package named IBM. The findings indicated that: the most important factors affecting positively (decreasing transaction cost value) on transaction cost value in construction projects in Egypt are: the financial position of the contractor, claims filed by the contractor, material substitutions by the contractor, contractor experience in similar type projects, contractor
s relationship with previous clients, contractor qualification to do the job, and procurement method. Transaction cost in the pre-contract phase of construction projects in Egypt is approximately (3–4) % of the contract value on average, and the average value for post-contract transaction cost is approximately (8–9) % of the contract value. All factors accredited as influential factors were taken into account and developed in an account model, the study devolved accounting models that can predict transaction cost value as a percentage of the project capital for construction projects in Egypt, two techniques were used, first is regression analysis with two methods (enter-stepwise), and second is the neural network, all models showed a high accuracy up to 95% in predicting transaction cost value for construction projects in Egypt.