In the study, it is aimed to examine the relationship between science fiction self-efficacy and spatial ability of science teacher candidates. The study is quantitative research and correlation research was used as a research design. The study group consisting of 200 science teacher candidates was formed by using the convenience sampling method. The data collection tools are Santa Barbara Solids Test, Spatial Ability Self-Report Scale and Science Fiction Self-Efficacy Scale for Science Teacher Candidates. Correlation analysis was performed in the analysis of data and Pearson Correlation for normally distributed data and Spearman Correlation for non-normally distributed data was used. As a result of analyzes, for science fiction candidates it was concluded that there is a moderately positive and significant relationship between spatial ability and spatial ability self-report, a moderately positive significant relationship between science fiction self-efficacy and spatial ability self-report, and a weak positive significant relationship between science fiction self-efficacy and spatial abilities. According to the findings, it was concluded that science teacher candidates make a consistent self-assessment in terms of their spatial ability. Moreover, it has been determined that some improvement in spatial ability can be achieved by using science fiction as a tool to develop spatial ability.