As a result of decreasing competition within the procurement of large construction projects in the Dutch construction sector, the need arises to understand the mechanisms behind the competition. Especially how decisions by clients on the procurement of construction projects impact the number of competitors. When this is understood, clients can contribute to solving the problems in the construction market. In an attempt to supply clients with insights to impact the number of competitors, this thesis tries to answer the main research question "To what extent do client decisions in the procurement of construction projects in the Dutch construction sector affect the number of competitors?". In an attempt to answer the question, it has been tried to create a model that predicts the number of competitors in the tender phase of construction projects. To be able to make predictions, data on factors that potentially affect the number of competitors is inserted in a regression analysis. Values are included that represent the variables project type, project value, project duration, client, contract type, market conditions, need for work, and tender method. The regression coefficients determined in this analysis can reveal how individual variables affect the predicted number of competitors. But the usability of these coefficients depends on the performance of the analysis set-up, i.e. the used regression model, chosen variables, and input data. It turned out that the achieved prediction performances were unsatisfactory. The best-performing model reached a prediction accuracy of 44.9%, just 8.7% better than a constant guess. As a result, it is concluded that the relationship between the client factors and the number of competitors could not be quantified. Additional research on the predictability within the data affirmed that it would be hard to make any predictions. The Entropy and Mutual Information within the data, identifying the predictability, showed that the independent variables in the data could only solve approximately 30% of the 'chaos' within the distribution of the number of competitors. The unsatisfactory model accuracies are an interesting result, given the fact that, for almost every factor considered influential on the number of competitors, a value has been included in the calculations. The question arises whether even a relation exists between client factors and the number of competitors. On top of that, the identified marginal relations pointed at the market conditions as most influential to the number of competitors. Further indicating that clients have a marginal effect on the competition within tenders. Therefore, it seems that clients are unable to solve the decreasing number of competitors in the construction industry. In order to show the absence of a relation between the number of contractors and client factors, further research is necessary. It is yet impossible to prove this absence because improvements can still be made to the utilized data and methodology.