Well-developed risk management tools provide critical support for successful delivery of construction projects. Considerable research has been conducted towards integration of risk management in front-end planning and in execution phases of this class of projects. The accuracy of these tools relies heavily on their respective assumptions and on the data used in their application. Consideration of risk in these tools utilizes two types of data: actual past records and estimated future data related to completion of projects under consideration.  The literature reveals that most published work in this area utilized these data either in bidding phase or in one of individual project execution phases to minimize the negative impact of risk on project cost and duration at completion. However, there is a lack of a comprehensive framework that employs both types of data in different phases of construction projects. This prevents construction practitioners from implementing an efficient risk management program. In this research, a new risk-based framework is developed, addressing limitations of existing models for different management functions over project lifecycle. The developed framework employs past performance data of construction organizations and projects in the bidding phase for risk maturity evaluation, contingency estimation, markup estimation, planning and scheduling, and progress reporting. The framework has five developed models. The first introduces a decision support model for risk maturity evaluation of construction organizations to identify their strengths and weaknesses in risk management processes, employing the Analytic Network Process (ANP) and fuzzy set theory. It enables construction organizations to assess and continuously improve their risk management capabilities. The second model introduces a new cost contingency estimation model considering correlations among project cost items, subjectively and objectively. It is also capable of modeling project cost contingency with and without the use of Monte Carlo simulation, which is deemed particularly useful when using subjective correlations. The third model introduces new pattern recognition techniques for estimating project markup. It utilizes Multiple Regression (MR), Artificial Neural Network (ANN) and Adaptive Nero-Fuzzy Inference System (ANFIS) techniques for that purpose, considering five factors: need for work, job uncertainty, job complexity, market condition, and owner capability. The fourth model introduces a newly developed multi-objective optimization model for scheduling of repetitive projects under uncertainty. The model considers the estimated cost contingency and the project markup in the total project cost and conducts, simultaneously, trade-offs between project duration, project cost, crew work interruptions, and interruption costs. It safeguards against assignment of unnecessary costly resources and provides a reliable project baseline. The fifth model presents a newly developed risk-based earned duration management model (RBEDM) that utilizes the generated project baseline in forecasting project duration at completion, considering critical activities only and their associated risk factors. It introduces a new risk adjustment factor (RAFcr) that quantifies the impact of future uncertainties associated with critical activities in estimating project duration at completion. This unique aspect of the developed model addresses the main drawback of earned duration management (EDM) its reliance on past performance data only. It also assists project managers in estimating more accurate and realistic required time to project completion.