Abstract: The function of pattern recognition of the generalised grey target decision method for mixed attributes is investigated. Thus, the theory framework of pattern recognition is proposed for the method which has its characteristics: variable target centre index, open and extensible decision-making basis (target centre distance). The theory framework goes as follows. First, the characteristics of standard patterns (regarded as the target centre index vectors) and the researched objects are determined; second, all mixed attribute values are converted into binary connection numbers (viewed as the index vectors); third, the generalised target centre distances between the researched objects and the target centre indices are calculated; and finally, the researched objects are recognised as some special patterns relying on the comprehensive weighted proximities. The application shows that the two investment enterprises y₁ and y₂ could select their suitable investment patterns S₃ and S₂ respectively (corresponding to their comprehensive proximities as 0.4270 and 0.2375 respectively).