Abstract Introduction The most used material for the preparation of the baseplates is the acrylic resin, but it can present distortions. Objective To evaluate preparation technique, region and storage time that presents less maladaptation of the base when made with self-cured acrylic resin. Material and method Models were created in gypsum type III simulating edentulous maxilla, as divided into 3 groups (n = 10): GC (control group) thermopolymerizable acrylic resin; G1 - manual adaptation technique and G2 - drip technique. For the measurements, silicone by condensation of light consistency that was interposed between base and model was used. With a hydraulic press, 50 kg pressure was applied leading the base of the model. The obtained mold was measured in the palate, canine and molar regions with a digital caliper at the following times: immediately after the base polymerization, at 24, 48, 72, 96 hours and one week. The results were submitted to statistical analysis. Result G1 presented maladaptation of 0.43 mm ± 0.10, while G2 obtained 0.39 mm ± 0.11. The lowest maladaptation occurred in the CG. The palate region presented greater maladaptation (0.52 ± 0.07) and the canine region, the lowest (CD = 0.27 mm ± 0.07 and CE = 0.27 ± 0.09); There was no statistically significant difference for storage times. Conclusion G2 presented lower values than G1, with no statistically significant difference. The palate region presented greater maladaptation, followed by molars and canines. The bases continued to maladaptation the model after the immediate polymerization, with no statistically significant difference.