Most of the complex systems we are considering in science, technology, social care and economy have uncertain and incomplete information concerning the system behaviour, its structure, boundaries, interaction with environment, etc. In order to omit these troubles and information lack, we use sometimes statistics and probabilistic approach, fuzzy and rough sets methodology. As it is shown in this review paper, we can use with much success new methodology – Grey Systems Theory (GST), which do not need any assumption concern-ing the distribution of sample, and high amount of data, because minimal number of obser-vations for GST use is only n ≥ 4. Using GST one can forecast the future behaviour of complex system, evaluate interdependence of its observation vectors (cause and effect), and evaluate optimal decisions possible to undertake in a given situations of decision making, as well as clustering of the similar systems.