This thesis provides insights and recommendations for the identification and response to creeping crises during the early stages of construction projects, with a special focus on engineering consultancy firms such as Witteveen+Bos. While existing research on crisis management in construction projects focuses on the execution phase, there is still a gap regarding early-stage crises, particularly creeping crises that are crises that develop gradually. These crises are difficult to detect, often going unnoticed until they reach the tipping point, which is the moment the crisis is visible and it has already damaged the outcome of the project. The main research question of this study is: “How can engineering firms identify and implement early warning signals to detect and manage creeping crises in construction projects?" and it is divided into four sub-questions in order to structure the study in a more systematic manner. To answer the research question and to develop an actionable framework, a qualitative mixed-methods research design is chosen. The research includes a literature review spanning general crisis theories, High Reliability Organizations (HRO), sensemaking, mindfulness, and early warning systems. The second phase involves semi-structured interviews with Witteveen+Bos professionals, where the main objective is to get insights of the most common creeping crises within the organization, barriers to detection, and the existing tools and techniques. The final phase consists of a validation workshop with Witteveen+Bos’ professionals to assess the framework and get a final round of feedback. After carrying out an analysis, findings reveal six common creeping crisis types: scope creep under stakeholder pressure, effort compensation due to overcommitment, stakeholder misalignment, legal or regulatory infeasibility and obstacles, and external contextual drift. Based on these crises and the barriers encountered, a four-stage Early Warning System (EWS) framework is developed: signal detection, signal interpretation, coordinated response, and learning and system adaptation. This framework integrates HRO principles, mindfulness, and sensemaking mechanisms but is adapted into the engineering context, emphasizing soft skills, technical tools, and qualitative judgment. Additionally, the proposed framework is designed to be flexible and customizable, allowing adaptation to different crises, and project teams, with the option of choosing different tools and techniques, including qualitative mechanisms. The validation workshop confirmed the relevance and feasability of the framework. The participants reacted positive both verbal and nonverbal, and showed a strong recognition of the findings presented. Although feedback was provided, the four-phase framework was generally well understood as well. Ultimately, this research contributes a practice-oriented model that supports engineering firms in detecting and managing creeping crises proactively, embedding early crisis recognition within professional routines while enabling future integration of tools such as AI or predictive analytics.