Start-up subsidies are a frequently employed policy instrument, the use of which is justified by alleged market failure resulting from positive external effects and capital market imperfections. This article investigates whether the allocation of subsidies reflects a policy focus on addressing these market failure occurrences. However, using survey data from the East German state of Thuringia, logistic regressions reveal a rather random subsidization of start-ups. Furthermore, propensity score matching suggests that subsidized start-ups would have survived and thrived in any case, an indication of deadweight losses of start-up subsidies. The analysis points to serious information problems arising when subsidies should be allocated to remedy market failure. Making the situation even more problematic is that failure to precisely target start-up subsidies is likely to result in market distortions and ineffectiveness.