In this retrospective study, we evaluated the short- and long-term effects of slow maxillary expansion with Haas, hyrax, and quad-helix appliances on posterior crossbite (PXB) correction stability, and maxillary intermolar width and angulation, in the deciduous or early mixed dentition. The inclusion criteria were models and treatment notes of patients with PXB at the start of treatment (T1), after PXB correction (T2), and at least 2 years posttreatment (T3). Exclusion criteria were craniofacial anomalies, fixed appliance use, or more than 1 expander type. From 312 consecutive expansion patients, 74 Haas, 41 hyrax, and 45 quad-helix subjects were evaluated regarding PXB correction, intermolar width, and angulation and compared with published norms to separate treatment effects from growth. The mean ages at T1, T2, and T3 were 8, 9, and 13 years. There were no significant differences in PXB correction stability or treatment response at T2 and T3 among the 3 expanders. Expansion increased intermolar width by 5 mm and tipped each molar by 2.3 degrees. At least 2 years after expander removal, molar width decreased by 1.3 mm, and the molars uprighted by 6 degrees. Compared with noncrossbite norms, PXB subjects had narrower intermolar width before treatment and greater width after expansion, and were slightly wider at least 2 years posttreatment. Both younger age at T1 and retainer use resulted in statistically greater intermolar width at T3. Eighty-four percent of PXB correction remained with about one third of the initial expansion lost; retainer use and early treatment provided increased intermolar width. Haas, hyrax, and quad-helix appliances were equally effective. Slow maxillary expansion altered the PXB patients' maxillary widths from narrower to slightly wider than the widths of the noncrossbite norms.