Present study deals with the statistical analysis of long-term ground level ozone (O3) trend and the influence of meteorological variables on its variation over Delhi, India. Daily mean and maximum of O3 and meteorological data, obtained from India Meteorological Department, were arranged for the period of 9 years (1998–2006). Based on the preliminary correlation study of all the data with O3, six variables viz. daily maximum temperature, daily average relative humidity, dew point, wind speed, visibility, and total sunshine were selected. Classical additive time series decomposition technique was used to obtain seasonally adjusted long-term trend. To analyze the masking effect of meteorology, adjustment was made using Kolmogorov–Zurbenko filters followed by stepwise regression analysis to the smoothed series of O3 maximum and meteorological variables, which showed that long-term trend was independent of sunshine duration. Results indicate a significant increasing trend with annual increase of 1.13 % for O3 mean and 3 % for O3 maximum. Annual deseasonalized trend for seasonal cycle shows bimodal oscillations. About 43 % of O3 variation was explained by the selected meteorological factors and rest of variation attributed to factors like emission of precursor gases, pollutant transport, policy changes, etc. Among the three tested regression models, performance of Model 2 with variable temperature, wind speed, and visibility was found to be best that resulted in lowering of O3 trend. Large variability (23 %) was explained by the variable visibility depicted that the emission of primary pollutants not only provides the precursor gases but also control the local photochemical reactions.