In order to investigate the impact of pea canopy architecture and development on microclimate and infection by [i]Mycosphaerella pinodes[/i], two field experiments were conducted in 2009 and 2010 at Le Rheu (France) to obtain canopies contrasted in height, closure dynamic, leaf area index (LAI) and leaf area density (LAD). Three pea cultivars (Athos, Antares, Gregor) were sown at two (80 and 40 seeds/m(2) in 2009) and three densities (80, 40 and 30 seeds/m(2) in 2010) and microclimatic sensors were located inside the canopy (at the bottom and in the middle) and outside. Two main sources of wetness were identified: rainfall and dew. During rainfall periods, average daily leaf wetness duration (LWD) was about 15 h, and 3 to 10 h longer inside than outside the canopies. LWD was positively correlated with LAI until canopy closure during these periods. During dry periods when dew was the only source of leaf wetness, average daily LWD was short, decreasing as the canopy developed. Shorter LWDs were observed at the base than at the mid-level of the canopies and longer LWDs were observed outside the canopy and inside the less dense canopies irrespective of the cultivar. LWD was negatively correlated with canopy height and LAI during these periods. Slow wind speeds were recorded inside the canopies (less than 0.5 km/h) and no significant canopy effect was observed on air temperature. An infection model was developed and showed that only rainfall periods which induced long LWDs inside the canopy, were favourable to [i]M. pinodes[/i] infection under our climatic conditions and suggested a more favourable microclimate inside dense canopies.