Background: Studies that seek to link the impact of ozone with mortality display a number of uncertainties. This study sought to ascertain the functional relationship between ozone and mortality, and analyse its short-term impact on daily natural-, respiratory- and circulatory-cause mortality in Spain. Methods: We analysed data on daily natural- (ICD 10: A00-R99), respiratory- (ICD 10: J00-J99) and circulatory cause mortality (ICD 10: I00-I99) for each of Spain's 52 provinces across the period 2000-2009. We calculated the impact of ozone on mortality by reference to relative risks (RRs) using generalised linear models, and performed an overall meta-analysis. Attributable mortality was calculated on the basis of RRs. Results: A quadratic relationship was observed and the existence of a threshold value identified in 33 of the 46 monitoring stations judged to be valid for ozone data, this threshold value was located above the 80th percentile. The overall RRs for 10 mu g/m(3) increases in concentrations of the 03 threshold value were as follows: 1.033 (95%C1: 1.023-1.044) for natural causes; 1.089 (95%C1: 1.058-1.120) for respiratory causes; and 1.025 (95%CI: 1.018-1.033) for mortality circulatory causes. The annual total of natural-cause ozone-attributable deaths was 499 (95%CI: 277-717), 126 of which (95%CI: 54-194) were due to respiratory causes and 167 (95%CI: 39-293) to circulatory causes. Conclusions: High ozone concentrations display a quadratic relationship with daily mortality. While this effect is more pronounced in respiratory causes, there is also a short-term effect on natural- and circulatory-cause mortality.