In a randomized outreach program, households offered enrollment assistance joined a benefit program at a higher rate than households not assigned the offer. Some control households may have learned about the program from neighbors, and some offered households could not be reached.