The Constellation-X mission is a follow-on to the current Chandra and XMM missions. It will place in orbit an array of four X-ray telescopes that Will work in unison, having a substantial increase in effective area, energy resolution, and energy bandpass over current missions. To accomplish these ambitious increases new optics technologies must be exploited. The primary instrument for the mission is the Spectroscopy X-Ray Telescope (SXT), which covers the 0.21 to 10 keV band with a combination of two x-ray detectors: a reflection grating spectrometer with CCD readout and a microcalorimeter. Mission requirements are an effective area of 15,000 cm(2) near 1 keV and a 15 arc-sec (BPD) image resolution with a goal of 5 arc-sec. The Constellation-X SXT uses a segmented design with lightweight replicated optics. A technology development program is being pursued with the intent of demonstrating technical readiness prior to the program new start. Key elements of the program include the replication of the optical elements, assembly and alignment of the optics into a complete mirror assembly and demonstration of production techniques needed for fabrication of multiple units. These elements will be demonstrated in a series of engineering development and prototype optical assemblies which are increasingly flight-like. In this paper we present image angular resolution error budgets for the SXT and for the Optical Assembly Pathfinder #2 (OAP2), the first of engineering development units intended to be tested in x-rays. We describe OAP2 image error sources and performance analyses made to assess error sensitivities. Finally we present an overall prediction of as-tested imaging performance in the x-ray test facility.
