We present a detailed analysis of an explicit model of warped D-brane inflation, incorporating the effects of moduli stabilization. We consider the potential for D3-brane motion in a warped conifold background that includes fluxes and holomorphically embedded D7-branes involved in moduli stabilization. Although the D7-branes significantly modify the inflaton potential, they do not correct the quadratic term in the potential, and hence do not cause a uniform change in the slow roll parameter eta. Nevertheless, we present a simple example based on the Kuperstein embedding of D7-branes, z 1 = constant, in which the potential can be fine-tuned to be sufficiently flat for inflation. To derive this result, it is essential to incorporate the fact that the compactification volume changes slightly as the D3-brane moves. We stress that the compactification geometry dictates certain relationships among the parameters in the inflaton Lagrangian, and these microscopic constraints impose severe restrictions on the space of possible models. We note that the shape of the final inflaton potential differs from projections given in earlier studies: in configurations where inflation occurs, it does so near an inflection point. Finally, we comment on the difficulty of making precise cosmological predictions in this scenario. This is the companion paper to Baumann et al (2007 Phys.Rev.Lett.99 141601).