During the past few years, the author and co-workers have investigated various cases of distress occurring in post-tensioned structures. In most cases these distresses did not jeopardize the safety of the structure. However, they have led to unsightly cracking and leaks as well as general concern on the part of the owner. Most of these deficiencies could have been predicted by the design engineer. This paper is intended to alert the design profession to the cause of these problems in order to reduce their rate of occurrence in the future. The cases illustrated were derived from actual field investigations rather than from conceptual analysis. In general, difficulties occur with post-tensioned structures due to structural detailing and design. Investigations have revealed inadequate consideration of the following factors: creep of the concrete under stress; the void space of the tendon conduit, which is ineffective until after grouting; bending in the direction opposite the dead load bending, which results from the tendon stress; the buttressing effect of other portions of the structure; and the effects of the sequence of tensioning tendons. In example one, Slab 1 was cast and stressed. Slab 2 was then cast against slab 1 with provision at the joint for a groove to receive sealant. Within one year creep occurred in the slabs. To prevent opening of the joint, the post-tensioning strand should have continued through the construction joint, thus eliminating the dead end in slab 2. This would have allowed the slab to move one-half of an inch at its far end rather than at mid-slab, which would now be a fixed construction joint. In example two, the American Concrete Institute's minimum requirements for tendon spacing are based an aggregate size, which determines whether the concrete can surround the tendon conduit. In the case investigated, the load of the structure at the time of post-tensioning was approximately two-thirds of the total design load. In addition, the I-beam's tendon configuration was such that, in some cases, shears induced by tendon stresses added to the shears produced by the dead and live load. Therefore, as a result of the reduced web area the horizontal shear stress during tendon tensioning was approximately 2 and one-half times more than the anticipated unit horizontal shear stress of the total web. This condition could have been avoided by thickening the web of the member, or by adjusting the drape of the tendons to prevent all of them from interrupting- at a single plane the effectiveness of the web.