Development aid does not always exert the desired positive effect on economic growth in recipient countries and it is even feared that it may reduce total factor productivity (TFP) and may discourage recipient countries' efforts. This study seeks to contribute to the research on aid transmission channels, in particular on macroeconomic channels such as private investment, domestic savings and the real exchange rate. By using panel data from 27 recipient countries over a 25-year period (1985-2009) this study aims to analyze the impact of the different forms of aid (grants, loans, bilateral and multilateral) on productivity, controlling for institutional factors and economic policy, using time-series panel techniques and focusing solely on the aid-productivity link. In order to examine possible vicious circles of aid, we run quantile regressions to ascertain whether aid is less effective in countries from the lowest TFP quantiles. To check for TFP-impeding conditions that are supposedly present in those quantiles, threshold regressions are performed to detect the ineffectiveness of aid below certain thresholds, including those of institutional quality, investment-to-GDP ratio, or domestic savings-to-GDP ratio. We find differences between the impact of aid in the form of grants and loans and the impact of bilateral and multilateral aid, with evidence that aid reduces TFP growth in the 0.1 and 0.25 quantiles. The search for sensible threshold values of aid impeding factors (institutional quality or key macroeconomic variables) was without result.