Gas and volatile organic compounds (VOCs) release in soil is known to be linked to microbial activity and can differently affect the life of organisms in soil. Electronic noses (E-noses) are sensing devices composed of sensor arrays able to measure and monitor gases and VOCs in air. This is the first report on the use of such a sensing device to measure specifically microbial activity in soil. In the present study, γ-irradiated sterilised soil was inoculated with Pseudomonas fluorescens. To be sure for a rapid microbial growth and activity, two pulses of nutrient solution with organic and inorganic C, N, P and S sources were added to soil and the resulting microcosms were incubated for 23 d. During the incubation, respiration and enzyme activities of acid phosphatase, β-glucosidase, fluorescein diacetate hydrolase and protease, were measured, and microbial growth as global biomass of vital cells based on substrate-induced respiration (SIR-C mic) and enumeration of viable and culturable cells by means of dilution plate counts (CFU) were also monitored. Concurrently, VOCs and/or gas evolution in the headspace of the soil microcosms were measured through the E-nose, upon their adsorption on quartz crystal microbalances (QCMs) comprising the sensory device. The E-nose typically generated an odorant image ( olfactory fingerprint) representative of the analysed samples (soils) and resulting from the concurrent perception of all or most of the analytes in headspace, as it commonly happens when several selective but not specific sensors are used together (array). The basic hypothesis of this study was that different soil ecosystems expressing distinct microbial metabolic activities, tested through respiration and enzyme activities, might generate different olfactory fingerprints in headspace. Furthermore, the possibility to detect several substances at the same time, released from the soil ecosystems, possibly deriving from both abiotic and biotic (microbial metabolism) processes provides an “odorant image” representative of the whole ecosystem under study. The E-nose here used succeeded in discriminating between inoculated and non-inoculated ecosystems and in distinguishing different metabolic and growth phases of the inoculated bacteria during incubation. Specifically, E-nose responses were proved highly and significantly correlated with all hydrolytic activities linked to the mobilisation of nutrients from soil organic matter and their cycling, with CO 2 fluxes (respiration and presumed heterotrophic fixation) and with P. fluorescens population dynamics during exponential, stationary and starvation phases measured by SIR-C mic and CFUs. Interestingly, the E-nose successfully detected soil microbial activity stimulated by nutrient supply, even though none of the catalytic activities tested directly produced VOCs and/or gases. The E-nose technology was then proved able to supply a real holistic image of microbial activity in the entire gnotobiotic and axenic soil ecosystems.