<p>Variable selection in functional linear regression is a critical concern when handling multiple functional predictors. Most existing methods rely on the least squares loss function, which makes them vulnerable to outliers or errors with heavy tails. It is highly desirable to adopt a robust variable selection procedure to address these challenges. In this article, we propose a robust variable selection method based on the group variance inflation factor (VIF) regression procedure. By utilizing the percentage of variance explained criterion, we initially apply functional principal components analysis to transform a multiple functional linear model into a finite grouped regression model. Subsequently, we adopt a robust version of the group VIF algorithm within this truncated model. We present a computationally efficient algorithm along with corresponding numerical experiments. Furthermore, a series of numerical experiments, including two real-world data examples, are conducted to evaluate and demonstrate the performance of our proposed methods relative to alternative group variable selection approaches.</p>