This work is devoted to solving the identifying software duplicates problem. To do this, a short survey of existing search methods is made, consisting of the following: textual, lexical, syntactic, metric, semantic. Then a comparative analysis of the methods is carried out according to the following criteria: accuracy, completeness, speed, resource efficiency, scope of implementation; the comparison results are given in tabular form. Also, promising approaches for searching for duplicates are considered, namely, the following: machine learning, graph analysis, syntax tree analysis, dynamic characteristics analysis, spatial characteristics analysis, abstract syntax analysis. Ways to continue investigation are indicated.