Inquiry-based instruction facilitates students' acquisition of scientific abilities, but little is known about how rapidly college students master these skills, particularly when this form of instruction includes scaffolding and especially when those scaffolds are removed. To address this gap, this study aimed to evaluate the effect that removing scaffolding from laboratory assignments had on students' application of scientific abilities in an inquiry-based environmental course, a strategy that also permitted an assessment of the pace at which students developed these skills. Results show that the rate at which students successfully apply scientific abilities depends on their complexity. For example, over 85% of the students rapidly mastered lower-order abilities. These abilities included data representation and identification of uncertainty and assumptions, among others. In contrast, by the end of the semester, fewer than 45% of the students successfully evaluated the effects of uncertainty and assumptions on their claims, distinguished hypotheses from predictions, or provided a theoretical framework for their arguments. This unsuccessful application of higher-order skills coincided with the removal of scaffolds. A dynamic system model permitted an evaluation of the observed learning process, revealing a good agreement between obtained results and modeled outcomes. This good agreement suggests that external factors (e.g., cognitive demand, domain-specific knowledge) potentially played a critical role in the mastery of scientific abilities.