This paper reports a duopoly market experiment that examines the effects of price complexity on market prices. In my experimental posted-offer markets, each seller offers an identical good to buyers with homogeneous preferences. First, sellers simultaneously decide on the price and the tariff structure of their good, then buyers make their choices. Each seller can choose to have a one-, two- or three-part tariff. The tariff structure affects neither the value nor the price of the good but influences buyers’ ability to calculate the good's price. The main results show that high-price sellers choose high complexity more often than low-price sellers if buyers are simulated in accordance with the bounded rationality model of Carlin (2009). However, the evidence for this effect is weaker if the buyers are human subjects. Importantly, prices are higher when the sellers can confuse buyers using price complexity than when sellers interact with perfectly rational robot buyers.