Construction labour productivity (CLP) is a key performance indicator for determining the success of construction undertakings, and notably affects the profitability of construction companies. To this effect, the construction industry and researchers have pursued better ways of addressing the CLP problem. The CLP problem is a very complex problem that includes one or a combination of processes of: identifying factors that can influence CLP, modeling construction processes to effectively predict CLP, and proposing mitigation measures for improvement of CLP. Despite ongoing efforts, properly addressing the CLP problem remains a challenge in both research and the construction industry, because the related processes entail simultaneously capturing: 1) complexity arising from the subjective nature of some variables affecting CLP, owing to the use of linguistic terms such as low temperature, or poor safety practices; 2) complexity arising from the dynamic nature of variables; 3) complexity arising from the emerging nature of some variables affecting CLP, such as crew motivation; 4) complexity arising from the causal relationships between factors affecting CLP – hereafter called situational/contextual variables – which are context dependent and vary across different situations in which tasks are performed; and 5) the inputs of multiple heterogenous experts involved in addressing the CLP problem (i.e., construction practitioners), whose inputs vary owing to their backgrounds, experience, and varying areas of expertise. This research provides a comprehensive state-of-the-art literature review and content analysis on the topic of system dynamics (SD) as a viable tool to capture the dynamic nature of system variables and their complex causal relationships for CLP modeling. Moreover, this research provides a fuzzy analytic hierarchy process–fuzzy decision making trial and evaluation laboratory (FAHP-FDEMATEL) method to capture causal relationships between crew motivation and situational/contextual variables affecting CLP. This research also provides a fuzzy system dynamics–fuzzy agent-based modeling (FSD-FABM) method to model CLP. The FAHP-FDEMATEL, and FSD-FABM methodologies proposed in this study are demonstrated and validated using a real industrial construction project in Alberta, Canada. This research also provides modeling frameworks that employ FSD-FABM with multi-criteria decision making (MCDM) and reinforcement learning (RL), which can be used to formulate CLP improvement strategies. These proposed frameworks on decision-making have also been validated using a case study on real construction projects. The main contributions of this research are: 1) providing a state-of-the-art on SD research; 2) providing a systematic and structured model for determining causal relationship mapping between factors affecting CLP via the proposed FAHP-FDEMATEL method; 3) proposing a novel hybrid FSD-FABM for capturing and assessing complexities arising from non-linear behaviors and dynamic causal interactions between multiple factors in modeling and predicting CLP; and 4) proposing novel FSD-FABM-MCDM and 5) proposing a RL–FSD-FABM decision making frameworks that can be used to propose productivity improvement strategies. The results of this study indicate that the FAHP-FDEMATEL model was capable of providing a systematic and structured method to map the causal relationship mapping between factors affecting CLP, while considering expert weights. Moreover, the proposed FSD-FABM in this study was capable of predicting CLP while considering the causal relationships between crews’ motivation and situational/contextual factors. In this regard, the proposed models (i.e., FAHP-FDEMATEL and FSD-FABM) and frameworks (i.e., RL-FSD-FABM and FSD-FABM-MCDM) can be used to provide solutions to the CLP problem.