The extremely small accelerations of objects required for the the onset of modified Newtonian dynamics, or MOND, makes testing the hypothesis in conventional terrestrial laboratories virtually impossible. This is due to the large background acceleration of Earth, which is transmitted to the acceleration of test objects within an apparatus. We show however, that it may be possible to test MOND-type hypotheses with experiments using a conventional apparatus capable of tracking very small accelerations of its components, but performed in locally inertial frames such as artificial satellites and other freely falling laboratories. For example, experiments involving an optical interferometer or a torsion balance in these laboratories would show nonlinear dynamics, and displacement amplitudes larger than expected. These experiments may also be able to test potential violations of the strong equivalence principle by MOND and to distinguish between its two possible interpretations (modified inertia and modified gravity).