We used a randomized control trial to measure how the free distribution of hybrid seeds and chemical fertilizers for maize production affected their adoption by small-scale farmers in the subsequent seasons. Information on their demand for the same inputs was collected through sales meetings which we organized in 2009 and 2011 where the inputs were actually sold. It revealed that the demand for the inputs of the free-input recipients was significantly higher in both 2009 and 2011 than that of non-recipients; that of the neighbors of the recipients fell in-between. The initial treatment assignment has a persistent influence on the farmers' demand over the two years whereas the difference between the free-input recipients and their neighbors has been reduced to some extent. The reduction of their gap in the application level of fertilizers is partly driven by social learning through information networks. However, there was no clear * I wish to thank Takashi Kurosaki, Keijiro Otsuka, and other participants of the TRIO conference in Tokyo held on March, 2013 for their excellent comments. I also thank George Sentumbwe, Geoffrey Kiguli, and other research team members of Makerere University who contributed to the surveys and workshops in the project. I am also grateful for funding from the Global COE project, sponsored by the Ministry of Education, Culture, Sports, Science, and Technology, Japan. All errors remain mine.