The detector system consists of two large area proportional detectors of the same size, each of which consists of twin main counters surrounded by built-in guard counters acting also in proportional region and in anti-coincidence with the main counters. To reduce secondary electrons and beta-rays from the subject, front guard counters are also arranged at the side of the entrance window relative to the main counters. Main counters are separated from guard counters by a multi-wire cathode to eliminate wall effect by reducing net wall area exposed to the main counters. Main counters, which are provided with field tubes in order to correct electric field distortion near the end of the counters, exhibit excellent uniformity of efficiency distribution and gas gain distribution over the entire volume of the sensitive region defined by the field tubes. As a consequence, good energy resolution is achieved, for example, 12.6% F.W.H.M. for 13.6 keV L-X ray of 239Pu. A multi-wire cathode results in imperfect electrostatic shielding, which causes electrostatic interaction between main counters and guard counters, such as variation of effective volume and gas gain depending on the anode potential of the guard counters. calculation of potential distribution near the cathode wire revealed that the interval of the cathode wire should be less than about 3 mm and some experimental results are also presented. The background of this detector with the gas filling of 90% argon plus 10% methane of 1 atm is 199 counts/min without anti-coincidence and 13.8 counts/min with anti-coincidence in an ordinary laboratory, and 73.0 counts/min without anti-coincidence and 1.46 counts/min with anti-coincidence inside an iron room. Preliminary calibration of the detector efficiency exhibits the possibility of detecting about half of the maximum permissible plutonium lung burden in 100 min counting time.