Infrared (IR) spectroscopy is one of the most powerful tools in life science. It delivers molecular information about structure and functionality in a non-invasive manner. However, its sensitivity and spatial resolution remain insufficient for interrogating single biomolecules. Surface and tip-enhanced spectroscopy exploit nanoscopic localization of both the probing light and the sample to address these shortcomings. Diffraction limited surface-enhanced methods offer high enhancement factors for ensembles of molecules even in aqueous environments but typically come along with high costs of production. Tip-enhanced methods provide even higher sensitivities down to hundreds of molecules and nanometric spatial resolution but are so far slow and require dry samples. I present in this work several routes towards single-molecule IR spectroscopy using both approaches. A cost effective and reproducible method for preparation of surface-enhanced infrared absorption spectroscopy substrates was developed. These resonant disc antenna arrays allowed the microspectroscopic characterization of sub-fmol (10^-15 mol or 10^9 molecules) of active membrane proteins. Their applicability to a variety of biologically important environments highlights their relevance for spectroscopy in life science. However, surface-enhanced techniques lack spatial resolution necessary for single-molecule detection and localization. Therefore, I have designed a scattering-type scanning near-field optical microscope (sSNOM) for IR nanoimaging and spectroscopy. A lateral resolution of 30 nm was achieved on protein loaded membrane patches pushing the sensitivity beyond zmol (10^-21 mol or 600 molecules). The imaging speed was improved by a factor of 20 compared to conventional setups enabling µs time-resolved studies on biomolecules. The obvious combination of resonant substrates and sSNOM yielded no appreciable enhancement of sensitivity and calls for alternative strategies. As an application to life science, whole cell nanoimaging and spectroscopy of the archeon Halobacterium salinarum was accomplished from which a homogeneous protein density within the cell wall could be inferred. Adapting a total internal reflection illumination scheme provided first experimental evidence towards sSNOM in aqueous environments. Those experiments lay the foundation for the analysis of complex membrane systems in living cells. The sSNOM setup was modified to record the locally deposited heat via the anomalous Nernst effect to expand the scope of tip-enhanced methods. The domain wall within a ferromagnetic micro device was localized as a proof of principle. This method cannot only be applied to antiferromagnetic systems but bears great potential for near-field IR spectroscopy. In conclusion, I believe that these results pave the way towards single-molecule IR spectroscopy by combining surface-, tip-enhancement and novel spectroscopic readouts.