This study tests the Involvement Load Hypothesis, proposed by Laufer and Hulstijn (2001), and constitutes a first attempt to explore the existence of levels of awareness in incidental vocabulary development. It investigates the effect of three input-oriented tasks with different degrees of involvement load on incidental Spanish L2 vocabulary development and text comprehension. In addition, it addresses levels of awareness as measured by think-aloud protocols, and the effect of word concreteness. Results show significant differences between conditions in both vocabulary development and levels of awareness, but not in the direction predicted by the Involvement Load Hypothesis.