Air pollution is a global problem. Particulate Matter (PM) of aerodynamic diameter smaller than 2.5 μm (known as PM₂.₅) and NO₂ are important classes of pollutants because of their size and emission sources and potential effects of exposure beyond 25 μm/m³ and 40 μm/m³ annual mean respectively. This thesis presents work that has been done to develop new and miniaturized/non intrusive (<1 cm³ in volume) sensors for monitoring both classes of pollutants. A review of the current landscape of both sensor types was carried out and the challenges identified. For PM, it is the price (>$300), size (smallest ones are several tens of cm³ in volume) and the accuracy (±10%) of the sensors that motivated the design, simulation and subsequent fabrication of the miniaturized device. It is shown that the capacitive-based sensor is easily miniaturizable and has sensitivity to single particles flowing at a distance of up to 18 μm above the electrode surface. This new sensor concept and its simulated multiphysics model is unique because it uses thermophoresis to separate particles of PM₂.₅ and PM₁₀ from a single airflow. For the NO₂ sensors, the availability of selective sensors that function in humid environments is a major need. Further, both sensor types need to be robust against interferent species and environmental variations. In this thesis, I present chemiresistors based on graphene/carboxymethyl cellulose (CMC) and carbon nanotube/CMC composites capable of sensing low, down to 20 ppm and 6 ppm, NO₂ concentration respectively. The new sensors show selectivity to NO₂ because of the selective oxidation of the composite component CMC salt by NO₂. Due to the Solubility of CMC in water and response of the sensor to ppm-level NO₂, a washable textile-based NO₂ sensor based on a reduced graphene oxide/MoS₂ composite material was developed. The sensor has selectivity to NO₂ and can detect ultra-low (100 ppb) NO₂ concentration levels in >60% humid air. It can also detect down to 20 ppb NO₂ in dry air. The next objective, beyond the scope of this work, is to integrate both PM₂.₅ and NO₂ detection and monitoring. Commercial exploitation of the technologies developed is now being explored through a University spin-out.