The objective of the case-control study presented here was to examine the risk factors for macrolide-resistant Streptococcus pneumoniae. As part of a 44-center U.S. surveillance study, 1,817 unique isolates of S. pneumoniae were collected from November 2002 through April 2003. Seventy-five randomly selected macrolide-resistant isolates (cases) were each matched with one susceptible control. Macrolide use in the 6 weeks prior to sample collection was reported for seven cases and one control. The final conditional logistic regression model identified two statistically significant variables: a history of alcohol abuse was protective, while macrolide use in the 6 weeks prior to sample collection was a significant risk factor for macrolide-resistant S. pneumoniae. Macrolide resistance was associated with use of any antibiotic during the prior 6 weeks, and was most strongly associated with previous macrolide use.