To identify the functional contributions of cooperations among transcription factors on regulatory DNA is critical for understanding transcription activation. But so far there is a great lack of effective identifying methods. Here we describe a novel strategy, based on comprehensively perturbed experiments and a computational model, to identify the cooperations among NF-kappaB (p65), CREB, and AP-1 in transcription activation of human cytomegalovirus major IE1 promoter/enhancer (MIEP). In this strategy, functional profiles of protein-MIEP association and RNA synthesis are achieved through comprehensively perturbing the association of p65, CREB or AP-1 with MIEP and then subjected to the computational model. Consequently, the 'real' cooperations contributing to MIEP activation are found to comprise five but not seven types of potential cooperations. Thus, our research provides a facile systematic approach to identifying the DNA-organized cooperations among transcription factors and understanding transcription activation.