Previous research suggests that learning from combined textual and pictorial representations is generally more effective than from text alone. However, when these elements are spatially separated, learners must mentally integrate them, increasing cognitive load. According to cognitive load theory, such split-attention formats are less effective than integrated formats, which cluster related information and reduce cognitive load. This study examined whether finger tracing enhances learning from spatially separated or integrated multimedia materials. 124 undergraduate students were randomly assigned to one of four conditions in a 2 (Tracing: yes vs. no) × 2 (Format: split-attention vs. integrated) design. Retention, comprehension, and cognitive load were assessed. Contrary to expectations, no significant effects of tracing or format were found. These null results suggest that tracing and split-attention effects may depend on factors such as content complexity or tracing fidelity. The findings refine our understanding of boundary conditions for instructional design and guide future research.