Abstract: Body Mass Index (BMI) is considered to be the most popular measure for overweight and obesity. Numerous studies of BMI are limited to compute and interpret different percentiles of BMI and do not account for many other covariates affecting BMI. Conventional regression methods are used for estimating how covariates are related to mean values of the dependent variable but in many situations, we are interested in quantiles rather than in mean values as in the case of BMI analysis. The present study addresses the same using median regression. Some important covariates such as gender, age, marital status, daily working hours, daily exercise routine and number of meat-eaten days per week are included in the study and found to be significant.