The study investigated the influence of a connected learning approach for explicit nature of science (NOS) instruction and explicit argumentation instruction, in comparison with a non-connected approach, on students' NOS conceptions and argumentation skills. Participants comprised 42 students enrolled in two sections of grade 10 that were randomly assigned (flip of a coin) into two groups: Connected (22 students) and non-connected (20 students). Participants in both groups were taught by the same teacher. The duration for the treatment was 2 months and one unit about plants was addressed. Participants in one group received explicit NOS and explicit argumentation instruction and learned about NOS and argumentation components as connected and at the same time, while participants in the other group received explicit NOS instruction and explicit argumentation instruction as non-connected and learned about them as separate. An open-ended questionnaire, together with individual semi-structured interviews, were administered at the beginning and conclusion of the study to assess participants' NOS conceptions and argumentation components in socioscientific contexts. The results showed that the connected group outperformed the non-connected group in generating arguments and counterarguments. As for NOS, the results revealed that both the connected and non-connected approaches showed a general enhancement of the conceptions of the NOS aspects. These results are interpreted in terms of the reflective judgment model, argumentation as a framework, an explicit approach, cognitive load theory, and curriculum coherence.