In this paper we investigate the relationship between per capita income and foreign aid for a panel of131 (alternatively 52) recipient countries over the period 1960 to 2006 by employing annual data and 5-year averages. Reliance on standard panel estimation techniques (such as 2-ways FE estimation, panel GMM and SUR estimation) points to some pitfalls (impossibility of possible cointegration between aid and growth, autocorrelation of the error terms, endogeneity of the variables) that must be dealt with panel time series techniques (such as panel unit root test, panel cointegration tests, a panel Granger causality test and panel dynamic feasible generalized least squares estimation (DFGLS). Estimations with DFGLS show that aid has an insignificant or a minute negative significant impact on per capita income. This result holds for countries with above- and below-average aid-to-GDP ratios, for countries with different levels of human development, with different income levels and from different regions of the world. It can be shown that by not controlling for autocorrelation, one erroneously attributes to aid a larger, significant negative impact on per capita income. We also find that aid has a significant positive (even though) small impact on investment, but a negative and significant impact on domestic savings (crowding out) and the real exchange rate (appreciation).