P-printable sets were defined by Hartmanis and Yesha and have been investigated by several researchers. The analogous notion of L-printable sets was defined by Fortnow et al; both P-printability and L-printability were shown to be related to notions of resource-bounded Kolmogorov complexity. NL-printability was defined by Jenner and Kirsig, but some basic questions regarding this notion were left open. In this paper we answer a question of Jenner and Kirsig by providing a machine-based characterization of the NL-printable sets.