In 2004, following a cluster of kala-azar cases in Chatrakhali, West Bengal, India, we screened and treated this endemic village for leishmaniasis infection. In 2005, following new reports of kala-azar, we screened the village again and conducted a retrospective cohort study (exposure period: August 2004 to July 2005). We defined an incident case of leishmaniasis as a new seropositive sample (>or=1:1600 dilution in a direct agglutination test) in a person seronegative in 2004. We obtained information about potential risk factors and calculated the relative risk (RR) of infection for exposure to these factors. One hundred and fifty (20%) of the 751 residents acquired leishmaniasis in 1 year. Factors associated with infection included residing in homes with mud walls (RR 4.3), dampness in the home (RR 2.5), proximity to bodies of water (RR 2.5) and livestock ownership (RR 2.4). Sleeping dressed (RR 0.4), or under a bed net (RR 0.5) or in a cot (RR 0.6) were associated with a lower risk. High rates of infection indicated that transmission persisted in this community. Poor housing conditions were associated with a higher risk, while personal protection measures against vectors were effective. Major housing improvement and personal protection efforts are needed to protect this vulnerable population from leishmaniasis.