In the present paper Poisson distribution of the tuberculosis population has been calculated and along with this expression has been obtained for the tuberculosis individuals as their average number. For obtaining such estimates the conventional method that has been used so far is Maximum Likelihood Principle but the problem that has been associated with this conventional principle is that this method any prior information that is available about the parameters of the study has not been taken into account by this method. This missing link has been accommodated by the Bayesian perspective and thus in a way obtains the estimators in such a unique aspect which consider the prior information about the estimator and refined the data through this information. In this paper Jeffrey’s non-informative priors and other two types of prior distribution has been considered and the corresponding estimates that has been obtained is calculated along with the standard error which itself is based on the assumption of squared loss error function. On the state wise and year wise data of the patients suffering from tuberculosis in India, this procedure has been applied. When the random variable that has been considered for the study is state, then Bayes estimate proves to be better than the Maximum Likelihood method and when the year is considered as random variable then out of the two Maximum Likelihood Method proves to better than Bayes.
