Background: Paracoccidioidomycosis (PCM) is Latin America's most prevalent systemic mycosis, carrying an important social burden. Its agent, Paracoccidioides brasiliensis, has rarely been identified in nature. Studies characterizing acute/subacute PCM incidence and their relationship with climate variables are not available. This work analysed a series of acute/subacute cases that occurred in the Botucatu area, São Paulo State, Brazil, from 1969 to 1999, as an outcome of weather variability.

Methods: Stepwise regression of annual data was applied to model incidence, calculated based on 91 cases, from lagged variables: antecedent precipitation, air temperature, soil water storage, absolute and relative air humidity, and Southern Oscillation Index (SOI).

Results: Multiple regression analyses resulted in a model, which explains 49% of the incidence variance, taking into account the absolute air humidity in the year of exposure, soil water storage and SOI of the previous 2 years.

Conclusions: The correlations may reflect enhanced fungal growth after increase in soil water storage in the longer term and greater spore release with increase in absolute air humidity in the short term.