This paper examines the relationship between three government support measures (debt moratorium, credit guarantee programs, and payroll subsidies) and the firm's payment behavior on loans in Colombia. To do so, we take advantage of the COVID-19 pandemic and use it as a case study. Using highly granular data at the bank-firm level and a difference-in-difference approach, we find that firms subject to debt reliefs and government guarantee programs experienced a lower probability of default while these policies were in force. Subsequently, once the programs ended, the dynamic of the payment behavior of these firms was similar to that of those untreated. On the contrary, payroll subsidies did not affect firms' payment behavior. Regarding the effect on banks' risk assessment, our results suggest that participation in relief programs provided banks with new information about debtors' risk, which could indicate unintended consequences of government support programs.