Many studies have analyzed different tooth demineralization models, which generate artificial incipient lesions; however, most of them are complex, slow, not clear and results could not be employed in both primary and permanent teeth because of chemical content differences among them. This study evaluates a demineralization model on primary enamel, under three incubation periods; quantifying artificial incipient lesions formation, and depth by micro‐CT, complementing with SEM for morphological characterization. Sixteen healthy human anterior primary teeth extracted for prolonged retention and orthopedic/orthodontic reasons were included in this study, previous informed consent. The sample was randomly assigned to four groups n = 4: G_Control, G_2D, G_4D, and G_7D. Micro‐CT and SEM were performed during two stages: before demineralization (BD) and after demineralization (AD). A t‐student test was carried out to determine differences among groups (p ≤ .05). No incipient lesions were observed in control group. Artificial lesion depth was similar among experimental groups; values were from 38.16 ± 05.40 μm to 42.61 ± 04.75 μm. An amount of 14 to 17 artificial incipient lesions were formed per group, the extension and distribution were different for each incubation period. Five erosive lesions were produced in G_7D. All experimental groups were able to form incipient artificial lesions in primary enamel. SEM characterization revealed more pronounced changes on the enamel surface, as the days of immersion in the demineralization solution increased. The 4‐day incubation period is the most recommended for the demineralization model, due to the formation of incipient lesions only and its extension, which facilitates their assessment.