The purpose of this research is to find out and compare thematic learning outcomes using the Structured Numbered Heads (SNH) and Numbered Heads Together (NHT) models for fourth-grade students at SD Negeri 42 Palembang. The type of research used is a quantitative research experiment. The number of samples in this study amounted to 44 people from 2 classes through the technique of simple random sampling. Data collection techniques using a questionnaire (questionnaire), interviews, and tests. The data analysis technique used is the t-test and questionnaire analysis. The results obtained in this study were no significant comparison between using the SNH and NHT models. It is based on criteria testing the t-test hypothesis obtained t = -1.51 and t = 1.68, then -t or t – 1.68 -1.51 which means accepted and rejected. It means the second model is just as good and thing This is seen from the average result of the class test experiment I and experiment II between classes SNH and NHT difference No significant. No way direct results This shows that between the second model This No there's better and use the second model it can be customized with condition participant educate.