A decision of choosing the right detector technology for third generation thermal imaging systems is directly derived from the expectations and the requirements of these systems. It is now evident that third gen. thermal imager will still need the higher resolution capabilities as well as capabilities in multispectral detection and polarization sensitivity. Four technologies candidates are analyzed; the field-proved HgCdTe (MCT), uncooled microbolometer technology, Antimonide based materials and quantum well infrared photodetectors (QWIP). Taking into accounts the risks, maturity and technologies gap of each technology, we claim that for non-strategic applications (not low background conditions), QWIP technology is the most favorite. The ternary and super lattice Antimonide based material group seems to be theoretically the best alternative, but are not recommended due to it maturity and the high risk involved in this technology. We anticipate large penetration of the uncooled detectors to the low-end and medium-end market. The HgCdTe will still be in progress due to the inertion of the large funding and the strategic importance of this detectors technology.