The aim of this study is to examine graduate theses including the experimental study on programming instruction in the subject of “education and training” in Turkey by using content analysis method. The research problem of this study is "how are the tendencies of graduate theses including experimental study on programming instruction?" In this context, it focuses on the following sub-problems: Regarding the graduate theses with experimental study on programming instruction in Turkey; What is the distribution by year? What is the distribution by thesis type? What is the distribution by research method? What is the distribution by sample type? What is the distribution by data collection tool? What is the distribution by programming tool is used? What is the distribution by programming language is used? What is the distribution by duration of the experiment? This study is planned as qualitative research and it is based on document analysis as data collection method. The thesis review form that prepared by researchers was used to collect the data. The content analysis was used to analyze of the data collected in the study. As a result of the searching in the Thesis Center of Council of Higher Education, 29 theses which contain at least one of the terms "programlama", "kodlama", "coding" and "programming" in the abstract of the study in the subject of "education and training " were selected and examined for the purpose of research. The graduate theses included in this study were coded based on the thesis review form. The findings of this study were displayed in graphs showing the frequency and percentage values. According to the findings obtained from the graduate theses by using the thesis review form, it was found out that; The number of master's theses is more than the number of doctoral dissertations. The number of theses related to programming instruction has increased in recent years. The most preferred research method is the mixed research method. The most used data collection tools are the scales and the success tests. In addition, many theses have used more than one data collection tool. The most commonly preferred scale type is the programming attitude scale. The most preferred sampling type is the undergraduate level. The most of the undergraduate students are enrolled in the department of computer education and instructional technology in the faculty of education. The most preferred programming language for programming instruction is the C# programming language. In some of the theses, the multimedia materials developed by the researcher are used to support programming instruction. The impact of the media on teaching programming was investigated in these theses. In some of the theses, the block based programming tools are used. In these theses the effect of the programming tool on teaching programming is examined. The most preferred visual programming tool is Scratch. The most of the experiments lasted between 4 and 6 weeks. According to the results of the research, suggestions related to programming instruction can be listed as following; The research can be conducted for pre-primary, primary and secondary school students about programming instruction. The research can be conducted for students in various undergraduate programs other than the department of computer education and instructional technology. The problem-solving and software development approaches can be used in the research for programming instruction. Instructional design models can be used for effective programming instruction. The methods about how programming should be taught can be investigated and compared.