This paper investigates Double Nominative Constructions in Korean. Two main approaches were suggested to account for DNCs in Korean: Indirect Case-Assignment hypothesis (Yoon 1991) and Direct Case-Assignment hypothesis (Maling and Kim 1992). However, these approaches only account for DNCs with an intransitive verb because there has been no suggestion for DNCs with a transitive verb so far. This paper suggests empirical data about DNCs with a transitive verb and suggests some differences between two NPs under DNCs with an intransitive verb and a transitive verb. In particular, I aim to explain the difference in NP constituency, depending on the verb types by introducing two diagnostics. In this paper, I present two different Agree systems to explain licensing nominative Case under DNCs with an intransitive verb and a transitive verb. First, following Yoon's (1991) approach, DNCs with a transitive verb can be explained by Indirect Case-Agree. Furthermore, I argue that DNCs with an intransitive verb can be accounted for by postulating Upward Multiple Agree. This concept is a combination of Upward Agree (Wurmbrand 2012a,b, 2014) and Multiple Agree (Hiraiwa 2014), allowing a probe to agree with goals multiply and upwardly.