Experiments investigating the process of macroscopic quantum tunneling in a moderately-damped, resistively shunted, Josephson junction are described, followed by a discussion of experiments performed on very small capacitance normal-metal tunnel junctions. The experiments on the resistively-shunted Josephson junction were designed to investigate a quantum process, that of the tunneling of the Josephson phase variable under a potential barrier, in a system in which dissipation plays a major role in the dynamics of motion. All the parameters of the junction were measured using the classical phenomena of thermal activation and resonant activation. Theoretical predictions are compared with the experimental results, showing good agreement with no adjustable parameters; the tunneling rate in the moderately damped (Q (almost equal to) 1) junction is seen to be reduced by a factor of 300 from that predicted for an undamped junction. The phase is seen to be a good quantum-mechanical variable. The experiments on small capacitance tunnel junctions extend the measurements on the larger-area Josephson junctions from the region in which the phase variable has a fairly well-defined value, i.e. its wavefunction has a narrow width, to the region where its value is almost completely unknown. The charge on the junction becomes well-defined and is predicted to quantize the current through the junction, giving rise to the Coulomb blockade at low bias. I present the first clear observation of the Coulomb blockade in single junctions. The electrical environment of the tunnel junction, however, strongly affects the behavior of the junction: higher resistance leads are observed to greatly sharpen the Coulomb blockade over that seen with lower resistance leads. I present theoretical descriptions of how the environment influences the junctions; comparisons with the experimental results are in reasonable agreement.