In this study the concentrations of PM10, PM2.5, PM1 and [222] Rn and meteorological variables (atmospheric pressure, air temperature, and relative humidity) were measured simultaneously to find association of particulate matters and Radon gas emissions using multivariate statistical methods. The data (from 1512 samples for PM and 196 samples for Radon measurement) have been collected in six medical treatment floors of Imam Khomeini Hospital in Tehran (Iran) from June 2014 to June 2015, seven days per each season. In this study, we conducted a time-series analysis to evaluate the effects of indoor or outdoor PM10, PM2.5, PM1 and Radon on hospital sections. During our study period, the PM10 and PM2.5, PM1 average concentration were 27.75, 20.05, 15.50 and varied between 7-49 mu g/m3, 6-37 mu g/m3 and 5-33 mu g/m3, respectively. The records showed that the average of Radon emissions in six floors of building were 2.8 Bq m(-3), 1.8 Bq m(-3), 2.8 Bq m(-3) 3.2 Bq m(-3), 1.2 Bq m(-3), 0.83 Bq m(-3) and 0.53 Bq m(-3) respectively. Multivariate Manova analysis for four variables (season, day, floor no., location) and through Pillai's Trace 'Wilks' Lambda 'Hotelling's Trace, 'Roy's Largest Root methods are used for providing table of Tests of Between-Subjects Effects for PM1, PM2.5, PM10 arrays. The results summarized meaningful deference between PM1, PM2.5 and PM10 for some effects. For evaluating of effects of air condition (Temperature, Pressure, and Relative Humidity) on PM concentrations we applied Stepwise Linear Regression (LR) and found that increasing of pressure and decreasing of temperature cause all PM increase but the temperature is sensitive more to PM1 and PM2.5 and pressure is sensitive just for PM10. We found that in the first and second half of year the Radon emissions at Nursery location had same effect but in Outdoor location and in cold seasons it has been increased. Finally, based on stepwise regression model, we can report that the concentration of [222] Rn and PM2.5 and ambient pressure showed reverse and direct correlation respectively. Linear model has more than 50 percent reliability in correlations but study should be continued by others to find more correlations between Radon and parameters. To our knowledge, this is the first study in Iran, or even in Asian developing countries, to report the effect of PM10,, PM2.5, PM1, and [222] Rn emission simultaneously on morbidity. Our findings also suggest that Radon could serve as a valuable air quality indicator that reflects the health risks of airborne particles.