We used a smartphone to construct three-dimensional (3D) models of keloids, then quantitatively simulate and evaluate these tissues. We uploaded smartphone photographs of 33 keloids on the chest, shoulder, neck, limbs, or abdomen of 28 patients. We used the parallel computing power of a graphics processing unit to calculate the spatial co-ordinates of each pixel in the cloud, then generated 3D models. We obtained the longest diameter, thickness, and volume of each keloid, then compared these data to findings obtained by traditional methods. Measurement repeatability was excellent: intraclass correlation coefficients were 0.998 for longest diameter, 0.978 for thickness, and 0.993 for volume. When measuring the longest diameter and volume, the results agreed with Vernier caliper measurements and with measurements obtained after the injection of water into the cavity. When measuring thickness, the findings were similar to those obtained by ultrasound. Bland-Altman analyses showed that the ratios of 95% confidence interval extremes were 3.03% for longest diameter, 3.03% for volume, and 6.06% for thickness. Smartphones were used to acquire data that was then employed to construct 3D models of keloids; these models yielded quantitative data with excellent reliability and validity. The smartphone can serve as an additional tool for keloid diagnosis and research, and will facilitate medical treatment over the internet. Copyright © 2020 Elsevier Ltd and ISBI. All rights reserved.