Mainly, the current study demonstrate the comparison of panel cointegration tests. This study comprised of comparative assessments of 24 tests and their comparison using stringency criterion. Among the compared tests, 21 tests are based on the null hypothesis of no cointegration, while 3 tests are based on the null hypothesis of cointegration. As far as novelty of the study is concerned, none of the existing studies have used this stringency criterion for the power comparison of panel cointegration tests. within the framework of null of no cointegration, this study evaluates the performance of 21 tests included; Residual based tests (parametric/nonparametric), Maximum Likelihood based tests, Fisher type tests, Average Weighted Symmetric test and Error Correction Based tests. The current study also extended toward comparison of power 3 panel cointegration tests having null hypothesis of cointegration. This study also evaluates the empirical size of the under consideration panel cointegration tests using asymptotic critical values and found that most of the tests are oversized. Therefore, size of all these tests have been controlled by using simulated critical values. The results of current study depicts that two residual based tests (PdP_V, PdPrho) and average weighted symmetric based test (PAWS) performed in paramount way throughout all time and cross sectional dimensions. current study also reveals that maximum likelihood based test LR and second generation test W_Gt perform worst in the current scenario. The LM_OLS test having null hypothesis of cointegration performed better than LM_DOLS and LM_FMOLS. Three better performing tests have reasonably high bootstrap powers based on Fisher hypothesis.