Data collected in a large epidemiologic study were analyzed to examine respiratory health effects of residential coal use in 7058 school children living in the four Chinese cities of Chongqing, Guangzhou, Lanzhou, and Wuhan. A Scenario Evaluation Approach was used to develop two exposure variables, heating coal smoke and cooking coal smoke. Estimated lifetime exposures to heating coal smoke and cooking coal smoke were both classified into four-level ordinal scales, as follows: no reported exposure (control); lightly exposed; moderately exposed; and heavily exposed. Zero-one dummy variables were constructed for each exposure level other than the control level (total six variables). These variables were entered into the analytical model. We tested for exposure-response relationships using logistic regression models, while controlling for other relevant covariates, including an indicator variable of ambient air pollution levels. We observed monotonic and positive exposure-response relationships of exposure to heating coal smoke with modeled odds ratios (ORs) of phlegm, cough with phlegm, and bronchitis. Other health outcomes were not associated with such exposure in a monotonic exposure-response pattern. However, ORs for cough, wheeze, and asthma were all higher in the exposed groups than in the control group. We observed no consistent associations between cooking coal smoke and the examined health outcomes. We conclude that exposure to heating coal smoke could have adverse effects on children's respiratory symptoms and illnesses in these four Chinese cities.