The majority of firms in most developing countries are informal. We conducted a field experiment in Sri Lanka which provided incentives for informal firms to formalize. Offering only information about the registration process and reimbursement for direct registration costs had no impact on formalization. Adding payments equivalent to one-half to one month's profits for the median firm leads to registration of around one-fifth of firms. A larger payment equivalent to two month's median profits induces half of the firms to register. Among the firms not registering after being offered this larger incentive, many faced issues related to ownership of land. Three follow-up surveys at 15 to 31 months after the intervention measure the impact formalizing has on these firms. Although mean profits increase, this appears largely due to the experiences of a few firms which grew rapidly, with most firms experiencing no increase in income as a result of formalizing. We also find little evidence for most of the channels through which formalization is hypothesized to benefit firms, although formalized firms do advertise more. Finally, formalizing is found to result in a large increase in trust in the state.