Abstract : The microbiological component of the aqueous phase of clouds collected from the puy de Dôme summit (1465 m a.s.l.) has been studied. We investigated 18 cloud events distributed along the year. We observed typical concentrations of 5 x 104 cells mL-1 for bacteria, and about 5 x 103 cells mL-1 for fungi and yeasts. Abiotic characteristics, like season or geographical origin of the sampled air masses, are closely related to the microbiological variations observed. Ocean is likely a major source of micro-organisms for cloud droplets, conversely to urban areas. Only a few of the total micro-organisms are retrieved by cultivation (about 10 % for fungi, and less than 1% for bacteria). However, as shown by ATP concentration measurements, they are almost all viable. Variations in the cultivable microorganisms composition have been detected, and seem to be a result of the evolution of sources like vegetation. Most of the isolates have physiological properties that can allow them to survive, maintain and even possibly develop in cloud water, like pigments and ability to grow at low temperature. They also can efficiently degrade the main organic compounds present in cloud water: formate, acetate, lactate, succinate, methanol and formaldehyde. This metabolic activity is even observed at 5°C and could thus significantly influence cloud water chemistry. In addition, transformations pathways due to microbial metabolisms are highly similar to those involving photochemistry, which is considered as the main catalyser of atmospheric chemistry. Furthermore, several strains have an important ice nucleation activity and could play a microphysical role in clouds. To conclude, the thesis suggests that micro-organisms could be involved in chemical and physical processes occurring in cloud water, and their presence may have to be considered to reach a better description of atmospheric chemistry.