Remote and virtual laboratories are increasingly prevalent alternatives to the face-to-face laboratory experience; however, the question of their learning outcomes is yet to be fully investigated. There are many presumptions regarding the effectiveness of these approaches; foremost amongst these assumptions is that the experience must be 'real' to be effective. Embedding reality into a remote or virtual laboratory can be an expensive and time-consuming task. Significant efforts have been expended to create 3D VRML models of laboratory equipment, allowing students to pan, zoom and tilt their perspective as they see fit. Multiple camera angles have been embedded into remote interfaces to provide an increased sense of 'realness'. This paper draws upon the literature in the field to show that the necessary threshold for reality varies depending upon how the students are interacting with the equipment. There is one threshold for when they first interact - the establishment reality - which allows the students to familiarise themselves with the laboratory equipment, and to build their mental model of the experience. There is, however, a second, lower, threshold - the maintenance reality - that is necessary for the students' ongoing operation of the equipment. Students' usage patterns rely upon a limited subset of the available functionality, focusing upon only some aspects of the reality that has been originally established. The two threshold model presented in this paper provides a new insight for the development of virtual laboratories in the future.