Grade 4 Kenyan children attending 10 randomly selected public primary schools in Nairobi (urban) and the Muranga District (rural) were surveyed to establish the prevalence of symptom markers of asthma and to assess the impact of urbanization. A respiratory health and home environment questionnaire was administered at school to parents or guardians. The questionnaire response rates were 94.2% (568/ 603) for Nairobi and 89.6% (604/674) for Muranga. The prevalence rates for asthma, defined as "attacks of shortness of breath with wheeze", were 9.5% for urban and 3.0% for rural children (odds ratio (OR) urban versus rural: 3.42; 95% confidence interval (CI): -1.96-5.91). This urban-rural gradient persisted after adjusting for urban-rural differences in host factors (including duration of breastfeeding and family history of asthma and/or allergy), but was largely explained by urban-rural differences in environmental factors, including indoor animals, sharing a bedroom with a smoker, parental education, house ventilation and exposure to motor vehicle fumes en route to school (adjusted OR: 1.59; 95% CI: 0.70-3.55). Similar results were obtained for all other symptoms. These findings confirm the clinical impression that asthma is an important illness in Kenya and underline the need for the further study of environmental factors amenable to intervention, particularly in urban areas.