This paper is concerned with the estimation and statistical tests of hypothesises around parameters of generalized modified Bessel linear regression model by Non-Bayesian (classical) techniques under the assumption that the shape parameters ( ) are known . We used the maximum likelihood method in estimation and using the likelihood ratio criterion for structure statistical hypothesis tests around parameters of the Regression model .Also we noted that probability distribution for estimator of the vector of parameters , is following the generalized multivariate modified Bessel distribution and the statistic test for the vector of parameters under the null hypothesis and alternative hypothesis are following the central F distribution and the non-central F-Bessel distribution respectively.
