Background and Purpose: Chemical hazard identification and human health assessments require identification of data from board categories of relevant information. Multiple evidence streams need to be systematically reviewed, critically assessed, and integrated. Global efforts to better characterize thyroid hormone pathways have resulted in a continuously expanding heterogenous literature base with over 1,000,000 articles. Systematic review methods used to conduct hazard or risk assessments face the challenge of needing to have a clearly defined question(s) or exposure-outcome scenario for the review to be effective. Herein we describe an evidence mapping workflow that uses both manual and computational elements to catalog information in the large, heterogenous literature set describing thyroid hormone (TH) pathways. The utility of the workflow is illustrated with a case study to explore multiple research objectives using a single thyroid pathway evidence map. A systematic mapping approach involving computational methods to inventory thyroid-related biological events is described. Methods: To develop the workflow, a series of piloting exercises were conducted to establish a series of document-level labels that were applied to each reviewed article. The labels described the study attributes and all measured biological events and were used to inform supervised machine learning (ML) algorithms. To illustrate use of the workflow, and pulling from a broad literature corpus, two molecular initiating event (MIE) categories (thyroid hormone (TH) serum distribution and membrane transport proteins) were investigated with the objectives of 1) identifying potential candidate reference chemicals with a range of potencies for test method verification, and 2) development of new test methods. Data labels were developed as DistillerSRTM question-answer sets allowing for computational automation through machine learning and natural language processing. Results: From a narrowed corpus of 1662 articles focused on thyroid hormone (TH) serum distribution and membrane transport proteins, titles/abstracts were labelled by humans and computers. Of these, more than 600 articles described administration of chemicals in mammalian in vivo or in vitro test systems and were further evaluated to identify potential reference chemicals that could be used for test guideline development. Mining at the full-text level identified assays that used immortalized cell lines or primary cells that may provide support for development into development of new methods to screen for perturbations of processes involving TH distribution and/or membrane transport proteins. The full text inventory of 436 articles was mined and 96 articles described TH membrane transport protein functional assays conducted in cell lines. A total of 13 articles described serum TH protein assays conducted in cell lines and 5 studies were identified as describing potential new assay systems. Conclusions: This work demonstrates that application of a comprehensive series of data labels can enhance the assessment of diverse and complex evidence, allowing assessment of multiple research objectives using a single evidence map. This systematic mapping approach will help promote transparency and increased efficiency when developing more focused research questions for human health assessments. The views expressed in this abstract are those of the authors and do not necessarily reflect the views or policies of the U.S. EPA.