OBJECTIVES: Potentially avoidable hospitalizations (PAH) are increasing in Spain and account for more than 3% of all the discharges. One of the hypothesis to explain PAH lies on the suboptimal management of chronic disease at ambulatory level of care. In this study we aim at explaining the association between compliance with preventive recommendations in primary care and observed rate of PAH. METHODS: Compliance with evidence-based preventive recommendations regarding diabetes mellitus (DM), chronic heart failure (CHF) and chronic obstructive pulmonary disease (COPD)  was calculated for the 1,350 general practitioners of the Basque Health Service and compared with the annual (2011) observed rate of PAH between their patients’ panel (mean panel≈1,500 adults; population covered=1,942,873). PAH was defined by a restrictive set of ICD-9 criteria following recommendations of the Spanish group of studies on Variability in Medical Practice. Multilevel poisson regression was used and age, sex, status socioeconomic (SSE) and hospital of reference were entered as covariates. RESULTS: After adjusting by all the covariates, higher compliance of recommendations such as periodical spirometry or assessment of the correct use of bronchodilator inhalers shows association with lower rates of PAH by COPD (p<0.05). Periodical foot care, oftalmological exam or assessment of the cardiovascular risk is non-significantly (0.15≥p≥0.05) but consistently associated to lower rates of PAH by DM. No link was observed between compliance or recommendations for CHF and PAH due to this cause. The independent effect of the SSE was strong in all the diseases (p<0.001) (higher rates as privation increases). Variations regarding the hospital of reference were also observed. CONCLUSIONS: Adequate compliance of evidence-based recommendations limits, modestly but consistently, the incidence of PAH due to DM or COPD. The SSE and the hospital of reference explain part of the observed variability in hospitalization by chronic disease.