Linear regression analysis of template aging in iris biometrics The aim of this work is to determine how vulnerable different iris coding methods are in relation to biometric template aging phenomenon. This is considered to be particularly important when the time lapse between gallery and probe samples extends significantly, to more than a few years. Our experiments employ iris aging analysis conducted using three different iris recognition algorithms and a database of 583 samples from 58 irises collected up to nine years apart. To determine the degradation rates of similarity scores with extending time lapse and also in relation to multiple image quality and geometrical factors of sample images, a linear regression analysis was performed. 29 regression models have been tested with both the time parameter and geometrical factors being statistically significant in every model. Quality measures that showed statistically significant influence on the predicted variable were, depending on the method, image sharpness and local contrast or their mutual relations. To our best knowledge, this is the first paper describing aging analysis using multiple regression models with data covering such a wide time period. Results presented suggest that template aging effect occurs in iris biometrics to a statistically significant extent. Image quality and geometrical factors may contribute to the degradation of similarity score. However, the estimate of time parameter showed statistical significance and similar value in each of the tested models. This reveals that the aging phenomenon may as well be unrelated to quality and geometrical measures of the image.