We describe progress evolving an important limit of binary orbits in general relativity, that of a stellar mass compact object gradually spiraling into a much larger, massive black hole. These systems are of great interest for gravitational wave observations. We have developed tools to compute for the first time the radiated fluxes of energy and angular momentum, as well as instantaneous snapshot waveforms, for generic geodesic orbits. For special classes of orbits, we compute the orbital evolution and waveforms for the complete inspiral by imposing global conservation of energy and angular momentum. For fully generic orbits, inspirals and waveforms can be obtained by augmenting our approach with a prescription for the self force in the adiabatic limit derived by Mino. The resulting waveforms should be sufficiently accurate to be used in future gravitational-wave searches. PACS numbers: 04.30.Db, 04.25.Nx, 95.30.Sf, 97.60.Lf The late dynamics of a merging compact binary remains one of the greatest challenges of general relativity (GR). GR doesn't have a "two body" problem so much as it has a "one spacetime" problem: one must use the Einstein field equations to find the dynamical spacetime describing a multibody system. Although numerical relativity has made great progress in recent years (e.g., [1] and references therein), most astrophysically relevant progress has come from identifying a small parameter which defines a perturbative expansion. One such approach is post-Newtonian (PN) theory (e.g., [2] and references therein) -essentially, an expansion in interaction potential GM/rc 2 . PN theory works very well when the bodies' separation r is large. As the bodies come close, the expansion must be iterated to high order (though it may be possible to modify the expansion to improve its convergence [3]).Strong field binaries can be modeled very accurately if one member is far more massive than the other. The spacetime is then that of the larger body plus a perturbation due to the smaller body, with the mass ratio acting as an expansion parameter. This limit is not just of formal interest: binaries consisting of "small" compact bodies (white dwarfs, neutron stars, or black holes with mass µ ∼ 1 − 100 M ⊙ ) captured onto highly eccentric orbits of massive black holes (M ∼ 10 5 − 10 7 M ⊙ ) are important targets for space-based gravitational-wave (GW) antennae, especially the LISA [4] mission. Such captures are estimated to occur with a rate density ∼ 10 ±1 Gpc −3 yr −1 . Convolving with LISA's planned sensitivity, one expects to measure dozens to thousands of events per year [5]. Precisely measuring GW phase as the smaller body spirals into the black hole (driven by GW backreaction) and fitting to detailed models will determine system parameters with extraordinary accuracy. For example, the hole's mass and spin should be determined to < ∼ 0.1% [6]. It should even be possible to "map" the black hole's spacetime, testing whether it satisfies GR's stringent requirements [7,8].Thanks to their extremal mass ratio, a f...