[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] Learning to use software programs using worked examples in screencast videos presents a classic split-attention problem that requires learners to mentally integrate information from the video with a target application. While much is known about sound screencast tutorial design, little is known about the features of the learning environment, such as monitor configuration, that may influence learning from this form of instruction. An experiment was conducted with 42 novice learners to fill this gap by comparing the effects of two common monitor configurations which split attention in different ways. In one condition, subjects split attention temporally by toggling back and forth between the video and target application on one monitor, while the other condition required subjects to split attention spatially by shifting their gaze between the video and target application displayed on two side-by-side monitors. Effects due to the monitor set-up were assessed for measures of cognitive load, instructional efficiency, and motivation using a 2x2 study design that controlled for task order. Results indicated that cognitive load as measured through task evoked pupil response was significantly higher, pless than .05, for groups with two monitors during both instruction and testing, even after controlling for working memory capacity. Analyses of three gaze-related eye-tracking metrics and NASA-TLX ratings did not indicate any differences in workload due to the experimental treatment; however, attentional patterns indicated by the eye-tracking data were shaped by two- and three-way interactions between working memory capacity and the experimental conditions. Measures of instructional efficiency indicated that the single monitor set-up was most efficient as learners attempted to transfer learning. While task efficacy and relevance ratings increased after training, monitor set-up did not affect post-training motivation.