In this dissertation, we describe a novel method that we call Diffusing Colloidal Probe Microscopy (DCPM), which integrates Total Internal Reflection Microscopy (TIRM) and Video Microscopy (VM) methods to monitor three dimensional trajectories in colloidal ensembles levitated above macroscopic surfaces. TIRM and VM are well established optical microscopy techniques for measuring normal and lateral colloidal excursions near macroscopic planar surfaces. The interactions between particle-particle and particle-substrate in colloidal interfacial systems are interpreted by statistical analyses from distributions of colloidal particles; dynamic properties of colloidal assembly are also determined from particle trajectories. Our studies show that DCPM is able to detect many particle-surface interactions simultaneously and provides an ensemble average measurement of particle-surface interactions on a homogeneous surface to allow direct comparison of distributed and average properties. A benefit of ensemble averaging of many particles is the diminished need for time averaging, which can produce orders of magnitude faster measurement times at higher interfacial particle concentrations. The statistical analyses (Ornstein- Zernike and three dimensional Monte Carlo analyses) are used to obtain particle-particle interactions from lateral distribution functions and to understand the role of nonuniformities in interfacial colloidal systems. An inconsistent finding is the observation of an anomalous long range particle-particle attraction and recovery of the expected DLVO particle-wall interactions for all concentrations examined. The possible influence of charge heterogeneity and particle size polydispersity on measured distribution functions is discussed in regard to inconsistent particle-wall and particle-particle potentials. In the final part of this research, the ability of DCPM is demonstrated to map potential energy landscapes on patterned surfaces by monitoring interactions between diffusing colloidal probes with Au pattern features. Absolute separation is obtained from theoretical fits to measured potential energy profiles and direct measurement by sticking silica colloids to Au surfaces via electrophoretic deposition. Initial results indicate that, as colloidal probe and pattern feature dimensions become comparable, measured potential energy profiles suffer some distortion due to the increased probability of probes interacting with surfaces at the edges of adjacent pattern features. Measurements of lateral diffusion via analysis of mean square displacements also indicated lateral diffusion coefficients in excellent agreement with rigorous theoretical predictions.