Background: A previous study suggested that U.S. students who lived in Mexico for 1 year had a risk of diarrhea intermediate between the rate for newly arrived U.S. students and Mexican students; however, the study was not controlled for changes of risky behavior over time. Methods: An analysis of acute diarrhea occurring among U.S. and Mexican student groups living in Guadalajara, Mexico was conducted to explore the association of diarrhea developing during selected 28-day periods with length of residence, season, and risk factors such as locations of food consumption, consumption of tap water, unsafe ice, alcohol, and antibiotics. Results: Compared to U.S. and Mexican student groups, newly arrived U.S. college students in July had the highest rate of diarrhea (55%), highest enteropathogen isolation rate (46%), and most consumption of alcohol and antibiotics; they also ate most frequently at restaurants and in Mexican family homes. Compared to a 34% rate of diarrhea among newly arrived U.S. medical students in August, the rate was only 6% among established medical students in January. This drop in attack rate was attended by less tap water and unsafe ice consumption by established students in January compared to the habits of newly arrived students in January or August when risky behavior was otherwise similar among these groups. The role of tap water and unsafe ice in the acquisition of wintertime diarrhea is further supported by the relatively high 29% rate of diarrhea among U.S. medical students newly arrived in January, who also consumed more tap water and ice than established students in January. Enterotoxigenic E. coli disease was observed only during the summer months; whereas, Campylobacter jejuni disease and disease associated with no detected pathogen were more common in winter. Conclusions: These data imply that wintertime diarrhea in Guadalajara is more likely than summertime diarrhea to be waterborne and to be caused by agents such as viruses or previously unrecognized enteropathogens that were not assessed in this study. Among Mexican students who had the highest frequency of tap water consumption and eating from vendors, the low rates of diarrhea in summer (3%) and winter (5%) imply that they enjoy a substantial degree of immunity. Further study will be necessary to understand the role that immunity (or its loss during summer vacation back in the United States) might play in additionally accounting for observed differences in attack rates between newly arrived students and those who had lived for a period of time in Mexico. (J Travel Med 2:6-10, 1995)