There is an urgent need for improved or conceptually new, ultrafast microscopy techniques which combine the best spatial and temporal resolutions possible. Techniques need to be developed which fuse nanometer-or atomic-scale imaging, such as scanning probe microscopy with ultrafast, femto-or attosecond time resolution, which is now commonplace in ultrafast science. A multitude of highly relevant, technological processes, such as the conversion of sunlight into electrical [1-3] or chemical energy [4-6], involves, on a microscopic level, the concerted motion of electrons and atomic nuclei. These processes are therefore governed by microscopic dynamics taking place on ultrashort and ultrasmall scales. Our ability to design and optimize such devices thus crucially depend on a detailed analysis and understanding of the spatio-temporal dynamics of charge, spin and atomic lattice excitations.A broad range of time-resolved microscopy techniques are currently under development or already in use. These techniques include X-ray diffraction/microscopy with large-scale free-electron lasers [7][8][9] or tabletop sources [10,11], ultrafast electron diffraction [12][13][14][15] or microscopy [16,17], time-resolved photoelectron emission microscopy [18-20] (Chapter 10), ultrafast, aperture-based [21][22][23][24][25] or aperture-less near-field scanning optical microscopy [26][27][28], and femtosecond scanning tunneling microscopy [29][30][31][32].So far, most of the techniques mentioned earlier provide either a limited temporal or spatial resolution, and the implementation of microscopy techniques with true nanometer and (sub-)femtosecond resolution [33] remains a formidable challenge. A fundamental problem in many of these approaches is the generation of sufficiently bright and temporally short light or electron pulses which are both coherent and energetically tunable [34].