The impact of some atmospheric pollutants (PM10, PM2.5, O3, NO2, NO, SO2, CO), humidity and temperature were studied on the performance of thoroughbred racehorses. The study included 162 official handicap races held in 2012 in Santiago de Chile, at distances of 1000, 1100 and 1200 m, on a track in good condition, with a layout that included a bend, during the summer and winter months. The environmental variables were measured at the time of the race and were obtained from a monitoring station located 470 m from the equestrian center. The environmental variables showed an autocorrelation of variables, so they were reduced using principal component analysis. Subsequently, the principal components were correlated with running speed using Pearson’s method. Totals of 60.17 and 23.29% of the total variability of the data was explained by principal components 1 and 2 (PC1 and PC2), respectively. PC1 was mainly determined by NO, NO2, and CO (loadings~0.90) and secondarily by PM10, PM2.5, and SO2 (loadings~0.6), with which the data showed inverse associations, while with temperature and O3 it showed direct associations (loadings~0.7). In addition, this component correlated negatively with running speed (r = −0.50), while PC2 was not associated with this variable. In conclusion, using the principal component analysis strategy, it was determined that running speed is affected by air pollutants.