Objectives: Ordinary regression model is basically built for conditional mean to show the relationship between mean and some explanatory variables through a statistical model. Some assumptions like normality should be defined with an acceptable degree of confidence to correctly make inference from this type of model. If a measure other than mean is interested in or these assumptions are not satisfied due to outliers, this model will not be useful. However, quantile regression is robust to the outliers and is able to build a model for any quantiles (quartles, deciles and percentiles). Method: In this paper the application of quantile regression is illustrated in the context of mental health data. Conclusion: In our study, quantile regression findings have shown different relationship of age to mental health for men and women whereas these could not be achieved by mean regression at all.