We examine a two country aid model with performance intensive aid. The aid budget is determined by a donor country legislature, but allocated by a donor agency in terms of a performance criterion of its choice. Five sources of slippage in policy delivery are introduced: the donor agency observes the performance of the aid recipient imperfectly; the donor agency and the aid recipient are subject to inefficiency; the aid recipient experiences corruption; and adjustment of aid recipient performance is costly. Incentives between the aid donor and the aid recipient are intentionally aligned – to explore the best-case for the policy intervention. Immediate implications are then that optimal level of effort and governance in the aid recipient increase under performance intensity of aid. Moreover, performance intensity of aid has fundraising effects. While the aid recipient also has an incentive to increase the measurement noise in governance, this incentive is weaker than that to raise true underlying governance, is weakest for the poorest aid recipients and provided that the donor agency is sufficiently efficient, measurement error itself will serve to raise optimal effort in the aid recipient. The importance of aid agency self-monitoring is thereby identified. Three important qualifications on the anticipated success of the policy emerge, however. Despite the elimination of incentive misalignment, aid donors may come to rely on performance intensity precisely where such a reliance is likely to be least successful in aid recipients. Second, performance intensity of aid maximizes its impact under conditions where it is supplemented by technical assistance to improve the effectiveness of own effort by the aid recipient. Finally, under convex adjustment costs the general class of optimal time paths in governance in aid recipients will be non-monotone, such that governance may get worse before they get better – optimally so. Performance intensive policy may thus appear to fail – immediately after imposition.