Cold air inhalation and exercise-induced bronchoconstriction (EIB) have both been used as measures of bronchial responsiveness. Both stimuli are often combined in the Nordic climate. The main objective of the present study was to investigate the climatic influence of cold temperatures upon exercise-induced asthma. The secondary aims were: (a) to assess metacholine bronchial hyper-responsiveness and EIB in children with bronchial asthma (n = 32; mean age 10.8 years) compared to children with other chronic lung diseases (CLD) (n = 26, mean age 10.1 years); and (b) to assess the influence of cold air inhalation upon EIB in the two groups of children. Methods used were: (a) the metacholine concentration causing a reduction in FEV1 of 20% (PC20-M), (b) maximum FEV1 fall (delta FEV1) after submaximal treadmill run (EIB test); and (c) delta FEV1 after submaximal treadmill run while inhaling cold (-20 degrees C) dry air (CA-EIB test). Geometric mean PC20-M did not differ significantly between the asthma children (1.28 mg ml-1) and the CLD children (2.90 mg ml-1). In the asthma children, mean delta FEV1 after EIB test was 12.8% vs 21.8% after adding cold air (P < 0.0001), compared to 5.2 and 7.4%, respectively (P = 0.03), in the CLD group. Maximum sensitivity and specificity for the EIB test were 69.8% at a fall in FEV1 of 6.8%; for the CA-EIB test, 72% at a fall in FEV1 of 10.2%; and for metacholine provocation, 56% at a PC20-M of 1.5 mg ml-1. In conclusion, children with bronchial asthma are substantially more sensitive to cold air than children with CLD, and EIB is markedly increased by cold air inhalation in asthmatic children, maintaining the specificity of the EIB test and increasing the sensitivity. The low sensitivity of the EIB test is probably influenced by the use of inhaled steroids. Metacholine inhalation test has less specificity and sensitivity in discriminating asthma from other chronic lung diseases.