As a rapidly expanding centre of government, trade, commerce and industry, Delhi, the Indian capital, presents an instructive location for studying the possible association between air pollution and adverse health effects. This study tries to determine the association, if any, between the air pollutants—sulphur dioxide, nitrogen dioxide, carbon monoxide, ozone, suspended particulate matter and respiratory suspended particulate matter—and daily variations in respiratory morbidity in Delhi during the years 2004–2005. Data analysis was based on the Generalized Additive Poisson regression model including a Lowess smoothing function for the entire patient population and subgroups defined by season. The best fitting lag period for each pollutant was found by testing its concentration at varying lags. The model demonstrated associations between daily visits and some of the pollutants (O3, NO2 and RSPM) but their strongest components were observed at varying lags. A single pollutant model showed that a 10 μg m−3 rise in pollutant level led to statistically significant relative risks (RR): 1.033 for O3, 1.004 for NO2, 1.006 for RSPM. The effect of particulate was relatively low, presumably because unlike other pollutants, particulate matter is not a single pollutant but rather a class of pollutants. This study, continued on a long term basis, can provide guidelines for anticipation/preparedness in the management of health care and hospital admissions.