In this paper, we explore statistical machine translation (SMT) approaches to automatic text simplification (ATS) for Spanish. First, we compare the performances of the standard phrase-based (PB) and hierarchical (HIERO) SMT models in this specific task. In both cases, we build two models, one using the TS corpus with “light” simplifications and the other using the TS corpus with “heavy” simplifications. Next, we compare the two best systems with the state-of-the-art text simplification system for Spanish (Simplext). Our results, based on an extensive human evaluation, show that the SMT-based systems perform equally as well as, or better than, Simplext, despite the very small datasets used for training and tuning.