Abstract: OBJECTIVES--To determine whether mean annual frequency and destination of equine travel was associated with exposure to Ehrlichia risticii and whether these associations were modified by horses' place of residence. DESIGN--Cross-sectional study. SAMPLE POPULATION--511 equine operations containing 2,587 horses were visited in New York state from a target population of 39,000 operations. PROCEDURE--Each horse was tested for serum antibodies against E risticii, using indirect fluorescent antibody. Information on the horse's travel history, farm's management practices, and surrounding ecology was obtained by personal interview and resource maps. Statistical analyses were performed on 2 cohorts of animals: all horses enrolled in the study and horses born on the property or that resided at least 4 years on the farm. Three county-based risk regions (RR) were identified by use of cluster analysis. RESULTS--Mean seroprevalence for each of the 3 RR was 2.4 (low risk), 8.5 (moderate risk), and 18.5% (high risk) for cohort 1 and 2.5, 8.0, and 18.4% for cohort 2. Among cohorts 1 and 2, pleasure riding and breeding trips were associated with exposure to E risticii, but horse residence (low, moderate, or high RR) was an effect modifier for these associations. Among cohort 1 and stratifying the analysis according to the RR for the travel destination, trail riding at low RR and trail riding at high RR were associated with exposure. Among cohort 2 and stratifying the analysis according to the RR for the travel destination, breeding trips were associated with exposure, and strong effect modification was present for horse residence (low, moderate, or high RR). CONCLUSIONS--Only certain types of travel to specific RR were associated with higher risk of exposure to E risticii. In many instances, travel was not associated, or was associated, with a reduced risk of exposure.