The mid-infrared (MIR) spectral range is of special interest for establishing optical chemical sensor technologies by allowing specific molecular identification and quantification, whether the sample is in a liquid, gas or solid form, in addition to providing highly sensitive, rapid, reagent-free and non-destructive detection. This thesis explores four different liquid- and gas- sensing applications and methods using MIR spectroscopy by integrating it with other technologies, such as microfluidics and fibre-optics. Firstly, fibre-optic integrated microfluidic devices were developed and tested for con- tinuous fluid monitoring. These showed good sensing capabilities for online, continuous and real-time liquid sensing in hard-to-reach locations. Next, this thesis presents the establishment and clinical testing of a novel method for continuous monitoring of the brain chemistry of traumatically brain-injured patients by MIR transmission spectroscopy. Here, the outlet of a cerebral microdialysis catheter is cou- pled to a micro flow-cell and the flowing microdialysate is continuously analysed. Clinical studies were carried out and showed the capability of this system for performing continuous patient monitoring over several hours. With further optimisation, the implementation of this system could lead to improved patient outcome. This thesis also presents a novel method and system based on MIR fibre-optic evanescent- wave spectroscopy, which enables enhanced detection of volatile organic compounds (VOCs). Here, a nanoporous silicon cladding was used to reversibly concentrate molecules close to the fibre surface, thus enhancing VOC detection. A significant increase in sen- sitivity was seen compared to that of an uncoated fibre and successful detection of three different VOCs, both independently and in binary mixtures, was achieved. Finally, this thesis introduces a simple and relatively low-cost fibre-optic sensor for in-line, real-time bioprocess monitoring. The sensor was successfully able to monitor varying concentrations of product (sophorolipids) in fermentation broth and was able to distinguish between the two types of generated product (acidic and lactonic sophorolipids). The work presented in this thesis showed that MIR-integrated sensors have great potential to provide novel and/or enhanced sensing solutions in a wide range of applications, including medical, industrial and environmental.