This work introduces a co evolutionary chromosome encoding scheme for evolving solutions in a high dimensional search space. The chromosome is divided in m “genes” and m different populations are created (one population per gene). Each one of the m populations evolves an specific gene and good references to genes in the remaining populations. The candidate solution is built using such references and the encoded gene. Individuals in the same population compete among them to find the best gene while individuals from different populations work together in order to find the best candidate solution. Finally, the best candidate solution is selected from all the populations based on its performance. Some experiments are conducted on well-known binary and real defined functions using three different evolutionary techniques. The obtained results indicate that the proposed approach is able to improve the underline evolutionary technique when evolving solutions for optimization problems in high dimensional spaces.