Antibiotic prescribing in primary care for common respiratory infections increased steadily until the mid 1990s, when the trend reversed noticeably. During the subsequent decade, antibiotic prescribing reduced by up to one-third in some countries. Explanations for this reduction have focused on a decline in the incidence and severity of common respiratory infections, and on the resulting decrease in the number of patients seeking consultation. We argue that evidence from primary-care research had a central role in changing the practice of antibiotic prescribing, and discuss the concern that has arisen among some physicians around this issue. Targeted reductions in antibiotic prescribing constitute a balancing act between individual and societal concerns, pitting the expected gains in preserving the usefulness of an antibiotic against any given reduction in use. There may be unintended consequences for decreasing antibiotic use beyond a certain point without adequate supporting evidence. A new approach to antibiotic prescribing requires comprehensive research to answer why change is necessary, and how that change can be safely implemented. Future policies must move beyond a "one size fits all" mindset if public and provider behaviours are expected to become more congruent with the growing research evidence.