Spreading out study opportunities over time improves the retention of verbal material compared to consecutive study, yet little is known about the influence of temporal spacing on orthographic learning specifically. The current study addressed four questions: (1) do readers' eye movements during orthographic learning differ under spaced and massed conditions? (2) is the spacing effect observed in offline post-tests? (3) can readers' eye movements during learning be linked with learning success in offline post-tests? (4) can E-Z Reader simulate the spacing effect during orthographic learning? Eighty adults silently read sentences containing novel words while their eye movements were monitored. Sentences were read four times; half of the items were spaced while half were massed. Participants completed a post-test assessing their written word form learning (orthographic choice or spelling). Simulations with E-Z Reader were used to interpret the human data. During orthographic learning, massed items had shorter total reading times than spaced items. A spacing advantage was noted in the offline post-tests. Longer fixations during learning were associated with higher response accuracy at post-test. Implementing a processing deadline enabled E-Z Reader to simulate participants' eye movements; simulations suggested that massed items may have received less attentional processing than spaced items during learning. Temporal spacing results in longer fixations during learning and better learning outcomes using offline tests. The combination of human eye movements and computational modeling provides useful insights into how reading and memory intersect and points to new directions for future research.