In most breeding programs, genetic diversity is considered a fundamental principle and condition for selection. In other words, understanding genetic diversity and estimating genetic parameters such as heritability, genetic correlation, and variance components of growth traits is a necessary condition for breeding. Restricted maximum likelihood (REML) is one of the important methods introduced for analyzing data, and it does not have the limitations of variance analysis for unbalanced and heterogeneous data, such as negative estimates, lack of uniqueness in estimates, lack of appropriate distributional properties, and lack of a useful method for comparing different types of estimates. This method maximizes the non-stationary part of the likelihood function and has the advantage of assigning degrees of freedom to the fixed part of the model, which leads to unbiasedness. It can also be implemented on data that have been affected by selection and deletion; While one of the assumptions of variance analysis is that the data obtained must be random. Other applications of this method compared to conventional methods for estimating variance components include directly providing genotypic correlations with greater accuracy. Despite the correct and more precise estimation of genetic parameters with these estimators, this method has been used to estimate heritability and genetic correlation in a limited number of domestic studies. In the present study, the estimation of genetic parameters including genetic correlation and heritability of the Amygdalus scoparia variety was carried out by evaluating seedlings obtained from seeds of two populations of this variety in Kerman province using the REML method, so that ultimately, by identifying the structure, diversity, and existing genetic potential, better decisions can be made for the management and improvement of this valuable variety and the production of more desirable seedlings in the study area.