Previous aid effectiveness research often claims that foreign aid has been unsuccessful in increasing growth rates. This result could be due either to methodological weaknesses, or to genuine aid efficiency losses. Empirically, the author finds that once the best available techniques are employed, aid has a significant but moderate average effect on growth over the period 1970-2001. A promising explanation of why the estimated returns to aid are not larger is that bilateral aid had no significant effect on growth during the Cold War (pre-1990), but had a significant and sizeable effect thereafter. In contrast, multilateral aid seems to have had sizeable and significant effects throughout. These results imply that the negative conclusions drawn by earlier research should be interpreted in their proper historical context, rather than as a necessary condemnation of current aid effectiveness.