In 1982 and 1983, a descriptive environmental survey was conducted in 317 households with newborn infants in rural Bilbeis, Egypt. The incidence of infant diarrhea in these households was ascertained by twice-weekly home visits for a 1-year period (1981–1982). Using univariate and multivanate analyses, the authors identified household factors that were statistically associated with infant diarrhea incidence, including number of children in the house under 4 years of age; number of persons per household; incidence of diarrhea in other family members; having a dirt (vs. concrete) dining room floor; having multiple living areas in the house; having a house or roof in need of repair, using well water rather than tap water for cooking or bathing; the absence of a sewer for waste bathwater; food being left out at room temperature between meals; and having many rodents in the house. Two practices involving interaction with the environment appeared to be protective: butchering of cattle by the family for home consumption, and protection of the infant from flies by a veil during napping. The combined household variables explained 25% of the variance in the total incidence of diarrhea. Categories of variables that accounted for most of the total variance explained by environmental factors are, in decreasing order: house structure (28%), water usage (24%); toilet and bathing area (12%); animal management (11%); food preparation area (10%); hygiene (8%); and wastewater management (6%). This approach may be useful in identifying environmental charactenstics whose change would reduce diarrheal illness among infants.