Current developments in optical microscopy aim to visualise complex dynamic biomolecular processes close to their native state. To capture transient phenomena, rapid three-dimensional stacks are acquired by translating the objective or sample stage to refocus into different depths of the specimen. Such conventional refocusing strategies introduce vibrational artefacts when imaging specimens that are in direct contact with the immersion media of the objective. Remote focusing is a methodology in which agitation-free refocusing can be performed using high numerical aperture (NA) objectives without compromising on resolution or imaging speed. It compensates for aberrations from the imaging objective by introducing equal and opposite aberration with a second microscope placed in reverse to the first. As the NA of the imaging objective increases, there are significant constraints placed on the tolerance in optical design to reach perfect phase-matching condition. In the first part of the thesis, the computational model developed to predict the performance of remote focusing microscopes is presented. From the model, the increased sensitivity of high-NA systems to magnification mismatch is inferred where the diffraction limited volume reduces by half for a 1% error. Informed by the sensitivity analysis, the decrease in resolution across depth for a remote focusing microscope with a 4% magnification mismatch is demonstrated. A protocol for magnification and resolution characterisation is presented and is applied to a novel Spinning Disk Remote Focusing microscope. The microscope is then applied to perform live volumetric imaging to study the normal neural activity of Platynereis dumerilii larvae. The studies presented here paves way for a standardised characterisation of remote focusing systems allowing for wider implementation. In the final part of the thesis, the spherical aberration generated by the correction collar on an immersion objective is exploited to compensate for residual spherical aberration in an ideal remote focusing system. The wavefront aberrations are measured using a Shack-Hartmann sensor and sub-resolution beads are imaged for point spread function measurements. Results from the Shack-Hartmann measurements show a 60% increase in axial range compensated for spherical aberration. In addition, the contribution of off-axis aberrations to the overall image quality at defocussed positions is explored further.Engineering and Physical Sciences Research Council (EPSRC)