Because the discovery of new antimicrobial agents cannot be expected in the near future, we will have to manage with the antimicrobials currently available at least for the next decade or two. Therefore, attempts to prevent development of antimicrobial resistance are of major importance. The relationship of local antimicrobial consumption and antimicrobial resistance has been shown in many hospital studies but not in the community, even though this is where most antibiotics are used. At the beginning of 1990s, erythromycin resistance in group A streptococci increased rapidly in Finland. The geographical variations found led to a nationwide study of the possible relation between local erythromycin consumption and variations in erythromycin resistance in the community. Erythromycin resistance was found to be significantly (P = 0.006) linked to local consumption of erythromycin. In further experiments, we found that a new erythromycin resistance phenotype belonging to the T4 serotype was spread over the whole country; 83% of the erythromycin-resistant isolates were of this new phenotype in 1994. In 1991, recommendations were given to reduce use of erythromycin in Finland. Following these recommendations, macrolide consumption decreased by 40% from 1991-1994. Studies are now in progress to evaluate the effect of this reduction on erythromycin resistance of group A streptococci.