Despite the severity and high frequency of cost overruns in construction projects, it is considered a rule rather than an exception. This phenomenon occurs when the initially estimated budget does not suffice the actual costs needed to realise the project. Uncertainty and bias challenge the ability to make cost estimates. Incorrect estimates of indirect costs, including risks, reservations and project organisational costs, are acknowledged as a frequent cause for cost overruns. Regardless of the risk and cost contingency management tools in literature and their application in practice, project uncertainties are still underestimated and insufficiently managed. During the execution phase projects are monitored with performance indicators, which provide information about past events or activities. This means these indicators are lagging, hence it limits proactive project management. On the contrary, leading indicators could function as early warning signs as they present themselves before potential issues arise, or risks materialize. The objective of this research is to improve the cost contingency process by implementing early warning signs, in order to proactively manage projects and minimise cost overruns and materialized risks. In order to achieve this objective the following main research question is formulated: “How can early warning signs be used in large construction projects to improve accuracy of cost contingency estimates and management in order to minimize cost overruns?” Concluding from the empirical study, early warning signs can be implemented in the cost contingency process by shifting the focus of risk management from mitigating and managing the effects of risks towards the causes of risks. Subsequently, risk strategies and mitigation measures can be more effective when aimed at the origin of the issue, instead of the direct issue leading to the risk. It is valuable to examine the risks as a system or network, instead of single problems, because it provides a better representation of reality and the links between risks or other events reveal information about correlation or side effects. Consequently, mitigation measures can increase in efficiency, when one cause, leading to multiple risks, is targeted. Furthermore, materialization of a risk is detected as an early warning sign which contains information about causes or circumstances that trigger other risks as well. Additionally, the increase of a RISMAN score indicates a change in project circumstances and therefore can be used as an early warning for materialization of risks. The justification for an increase in RISMAN score is diverse, and should thus be analysed to reveal the specific information needed to prevent materialization of risks. A common database contributes to the early warning sign approach as well, because it overcomes the in the case study identified ‘surveillance filter’, by improving collection and transfer of information. Tracking the cost contingency increases the accuracy of this estimate in general, because it provides insight into the development during project phases and the actual spending pattern. And when this info is stored in a database it allows referencing from previous projects, hence including an ‘outside view’.