Background: Unilateral posterior crossbite is a common malocclusion, and early treatment is recommended to enable normal growth. There are several possibilities regarding choice of appliances used for correcting this malocclusion; however, when treatment is financed by public funds the decision needs to be based not only on the effects but also on the effect in relation to the costs. Objectives: The aim was to perform a cost-effectiveness analysis comparing quad helix (QH) and rapid maxillary expanders (RME; hyrax-type) in children in the early mixed dentition. Material and methods: Seventy-two patients were randomized to treatment with either QH or RME, at two different centres. Data were collected from the patient's medical records regarding success rate, number of visits, total treatment time, emergency visits, and so forth, together with answers from patient questionnaires concerning absence from school and use of analgesics. A cost-effectiveness analysis with both an intention-to-treat (ITT) and a per-protocol approach was performed, as well as a deterministic sensitivity analysis. Results: The success rate, one year after the completion of the expansion, was equal between groups according to the ITT approach. From a healthcare perspective, the mean cost difference between RME and QH was <euro>32.05 in favour of QH (P = 0.583; NS). From a societal perspective, the mean cost difference was <euro>32.61 in favour of QH (P = 0.742; NS). The total appliance cost alone was higher in the RME group <euro>202.67 resp. <euro>155.58 in the QH group (P = 0.001). The probability of RME having a higher cost was 71% from a healthcare perspective and 62.7% from a societal perspective. The total treatment time was 97 days longer in the QH group. In the deterministic sensitivity analysis, when using a higher valuation of the children's educational loss, the QH becomes <euro>58 more costly than the RME. There was a statistically significant difference in chair time and visits between centres (P < 0.001). Conclusion: The difference in costs between RME and QH is not statistically significant, however, there is a slightly higher probability that RME is more expensive than QH with a mean cost of an additional <euro>32 per patient from a healthcare perspective. Different work procedures at different centres indicate that logistics around the patient's treatment is a more important aspect than appliance used to decrease the number of visits and save chair time and thereby also costs.