This study aims to investigate the trends and topics in Computer-Supported Collaborative Learning (CSCL) research papers from the emergence to the present. For this purpose, all documents on the Scopus database (5775 papers) were gathered and analyzed through bibliometric analysis and Latent Dirichlet Allocation (LDA) analysis, one of the topic modeling-based machine learning methods. The titles, abstracts, and keywords of the papers were included in the analysis. The bibliometric findings first indicated the annual distribution and subject areas in addition to the top productive sources, affiliations, countries, and authors of the CSCL papers from the emergence to the present. The findings from the LDA analysis revealed nine topics in CSCL. The high-density topics are (1) instructional design for CSCL, (2) designing learner groups and learning environments for CSCL activities, and (3) design and development for social constructivist learning. The medium-density topics are (4) developing social affordances in learning groups and CSCL environments, (5) constructivist approaches in CSCL contexts, and (6) social agency with CSCL technology. The low-density topics, focusing more on specific issues in CSCL, are (7) CSCL technologies, (8) online collaborative problem-solving, and (9) knowledge construction in CSCL. The publication accelerations of these topics were observed upward over time from the emergence of the field. These findings imply that instructional design, social issues, constructivism, and CSCL technologies are the common characteristics of the gathered topics, which makes many of them highly interrelated. The findings were discussed based on the relevant bibliometric and other review studies on CSCL.