The primary objective of this research was to compare procedure-based task performance using three common mobile screen sizes: ultra mobile personal computer (7 in./17.8 cm), personal data assistant (3.5 in./8.9 cm), and SmartPhone (2.8 in./7.1 cm). Subjects used these three screen sizes to view and execute a computer maintenance procedure. Results from 65 student participants indicated a significant difference in completion times between the three screen sizes [F(2, 120)=690, p less than 0.050], but no differences in subjectively assessed cognitive workload, errors or performance time. Competing memory strategies in procedure-based tasks were revealed through the access frequency and amount of time subjects spent reading the procedure before starting the experiment [F(2, 105)=25.17, p less than 0.001]. When using the 7 in./17.8 cm screen, users read more of the procedure before executing the procedure. In contrast, when using the smaller screen, users sampled the procedure in longer intervals during execution. (Contains 11 figures.)