Abstract: Direct decision (DD) is the most frequently-used blind equalization algorithm with advantages of fast convergence and small residual errors. Genetic blind equalization is an effective method for blind equalization, which makes use of a genetic algorithm to solve blind equalization problems and has better global optimization characteristics. By studying characteristics of the DD algorithm and genetic blind equalization, this paper provides a dual-mode genetic blind equalization algorithm by using error signals. The proposed method combines advantages of DD and genetic blind equalization, and significantly improves operation efficiency. It also enhances convergence performance and avoids a larger operand in the genetic blind equalization algorithm. Effectiveness of the algorithm is verified in computer simulation.