The relationship between air pollution and the daily number of contacts (i.e., telephone calls and home visits) with or at Copenhagen Emergency Medical Service for children with and without respiratory illnesses was studied during a 91-d period (i.e., January 14, 1991, to April 14, 1991). A total of 12,132 contacts occurred. Diagnoses, which were recorded on the invoices for 5,307 contacts, revealed that 3,974 contacts were the result of respiratory illnesses. Regression analysis was used to investigate the short-term relationship between pollutants (i.e., carbon monoxide, nitric oxide, nitrogen dioxide, NOx, sulfur dioxide, ozone, and black smoke), measured at monitoring stations, and both the number of all contacts for children and the number of contacts for children with respiratory illnesses. Temperature and systematic effects that were the result of holidays and weekends were controlled for, after which only nitric oxide and NOx were associated significantly with the number of contacts for children who had respiratory illnesses. Nitric oxide and NOx, as indicators of traffic pollution, appeared, at low levels, to slightly exacerbate respiratory illnesses among children.