Pattern printing techniques have advanced rapidly in the past decade, driven by their potential applications in printed electronics. Several printing techniques have realized printed features of 10 mu m or smaller, but unfortunately, they suffer from disadvantages that prevent their deployment in real applications; in particular, process throughput is a significant concern. Direct gravure printing is promising in this regard. Gravure printing delivers high throughput and has a proven history of being manufacturing worthy. Unfortunately, it suffers from scalability challenges because of limitations in roll manufacturing and limited understanding of the relevant printing mechanisms. Gravure printing involves interactions between the ink, the patterned cylinder master, the doctor blade that wipes excess ink, and the substrate to which the pattern is transferred. As gravure-printed features are scaled, the associated complexities are increased, and a detailed study of the various processes involved is lacking. In this work, we report on various gravure-related fluidic mechanisms using a novel highly scaled inverse direct gravure printer. The printer allows the overall pattern formation process to be studied in detail by separating the entire printing process into three sequential steps: filling, wiping, and transferring. We found that pattern formation by highly scaled gravure printing is governed by the wettability of the ink to the printing plate, doctor blade, and substrate. These individual functions are linked by the apparent capillary number (Ca); the printed volume fraction (phi(p)) of a feature can be constructed by incorporating these basis functions. By relating Ca and phi(p), an optimized operating point can be specified, and the associated limiting phenomena can be identified. We used this relationship to find the optimized ink viscosity and printing speed to achieve printed polymer lines and line spacings as small as 2 mu m at printing speeds as high as similar to 1 m/s.