As microtopography can influence the contact behavior of materials, it is of great significance to study the correlation between morphology characterization parameters and contact performance. In the light of complex relevance of parameters, a method for screening roughness parameters (RP) to characterize contact performance is constructed to get the maximum influence parameters on the contact stress (CS) and avoid the error of experiential selection. First, Pearson's coefficient and back propagation (BP) neural network are utilized to elaborate on correlation level between RP and CS and to build the regression model. Then global sensitivity analysis (Sobol) and local sensitivity analysis (MIV and Garson) are introduced to demonstrate RP quantitative influences on CS and select main RP for characterizing contact performance. The research shows (1) in the correlation analysis, RP with high correlation and noncollinearity on σmax are Sa, Sdq, S5p, Spk, and Svk; With regard to Mpmax and τmax, Sa, S5p, Sdq, and Vmp are on display, (2) RP importance sequence based on the results of correlation analysis is Sa, Spk, Sdq, Svk, S5p for σmax, and Sa, Vmp, Sdq, S5p for Mpmax and τmax, and (3) For the comprehensive main parameters model, RP for characterizing contact performance under the three contact stresses are Sa, Spk, and Vmp, belonging to height parameter, function parameter, and volume parameter, respectively. According to definition, all of them can significantly affect the stress concentration and distribution on contact surface of materials, which validates the rationality of the method.