We use a laboratory experiment to explore the effect of a change in pre-trade anonymity in a quote-driven dealer-to-customer market, organised as a request for quote (RFQ). We consider two treatments in which dealers interact with two types of customers (informed or uninformed). In the first treatment, there is no anonymity: dealers know whether the customer that sent them the request for quote is informed or uninformed. In the second treatment, there is complete anonymity: dealers do not know the type of customers they are interacting with. We find that anonymity improves price efficiency, whereas it does not adversely impact dealers’ trading profits. Our results contribute to the debate on transparency versus the adoption of anonymity in financial markets.