We are developing resonator-QWIPs for long wavelength detection. The cutoff wavelength is about 10.5 microns. R-QWIP pixels with 25 micron pitch were hybridized to fanout circuits for radiometric measurements. In one of the material designs, we adopted a coupled double-well in each QW period to broaden the material absorption bandwidth. Detector structure that contains 19 QW periods is labeled as DX1, and that contains 8 QWs are labeled as DX2. Both materials are moderately doped to 0.5E18/cm3 to obtain a proper balance between quantum efficiency and dark current. For the DX1 R-QWIPs, we observed a peak QE of 37% under positive substrate bias and 20% under negative bias, showing a large polarity asymmetry. On the other hand, for the DX2 R-QWIPs, we observed similar QE of 35% and 34% under the respective polarities. The detection bandwidths in all cases are approximately 2 microns. We attribute the QE asymmetry in DX1 to the highly localized resonant optical intensity distribution in the resonator volume and the nonlinear potential profile of the QW layers. Under positive bias, the high field domain in the QWIP layer coincides spatially with the high intensity region and thus it yields a higher QE. And this observed QE matches the optical absorption predicted by EM modeling. For DX2, since its layer thickness is thinner, all the QWs are within the high intensity region irrespective to bias polarity. The location of the high field domain thus becomes irrelevant. Because of their moderate doping density, the dark current is observed to be the same as the background photocurrent under F/2 optics and 300 K background at the operating temperature of 65 K. In addition to the fanout study, we have also fabricated focal plane arrays for imaging demonstration. Their performance will be presented.