Recent results in the assembly of DNA into structures and arrays with nanoscale features and patterns have opened the possibility of using DNA for sub-10 nm lithographic patterning of semiconductor devices. Super-resolution microscopy is being actively developed for DNA-based imaging and is compatible with inline optical metrology techniques for high volume manufacturing. Here, we combine DNA tile assembly with state-dependent super-resolution microscopy to introduce crystal-PAINT as a novel approach for metrology of DNA arrays. Using this approach, we demonstrate optical imaging and characterization of DNA arrays revealing grain boundaries and the temperature dependence of array quality. For finite arrays, analysis of crystal-PAINT images provides further quantitative information of array properties. This metrology approach enables defect detection and classification and facilitates statistical analysis of self-assembled DNA nanostructures.