This purpose is to find out whether the mass media coverage and the attitudes of the LGBT community affect LGBT self esteem in Medan. The research method is quantitative descriptive correlative type with correlative study patterns by placing research variables in two groups, namely independent variables and dependent variables. The study population was the LGBT community in the city of Medan. Sampling uses Proportional Random Sampling technique with a selected sample of 67 people, the data collection method uses the scale of mass media coverage, the scale of community attitudes and the scale of self esteem of the LGBT community. Analysis of research data shows that (1) there is a very significant positive effect between the scale of mass media coverage and the self esteem of the LGBT community, which is indicated by the coefficient rx1y = 0.380 and p 0.01; (2) there is a significant positive effect between community attitudes and influencing community self esteem as indicated by the coefficient rx2y = 0.678 and p 0.01; for the two hypotheses above, the product moment analysis technique is used; (3) there is a significant influence between the mass media coverage and community attitudes with the self-esteem of the LGBT community in the city of Medan as indicated by the coefficient F = 27,341 and the correlation of the three variables namely R = 0.679 while the magnitude of the third contributors R2 = 0.461 with p 0.01 with the meaning of the magnitude of influence is 46.1%. The third hypothesis in this research is to determine the power of mass media coverage and community attitudes towards the emergence of community self esteem, the analysis used is Multiple Regression Analysis.