Together with the growing availability of data from electronic records from healthcare providers and healthcare systems, an assessment of associations between different environmental parameters (e.g., pollution levels and meteorological data) and hospitalizations, morbidity, and mortality has become possible. This study aimed to assess the association of air pollution and hospitalizations using a large database comprising almost all hospitalizations in Poland. This time-series analysis has been conducted in five cities in Poland (Warsaw, Białystok, Bielsko-Biała, Kraków, Gdańsk) over a period of almost 4 years (2014–2017, 1255 days), covering more than 20 million of hospitalizations. The hospitalizations have been extracted from the National Health Fund registries as daily summaries. Correlation analysis and distributed lag nonlinear models have been used to investigate for statistically relevant associations of air pollutants on hospitalizations, trying by various methods to minimize potential bias from atmospheric parameters, days of the week, bank holidays, etc. A statistically significant increase of respiratory disease hospitalizations has been detected after peaks of particulate matter concentrations (particularly PM
2.5
, between 0.9 and 4.5% increase per 10 units of pollutant increase, and PM
10
, between 0.9 and 3.5% per 10 units of pollutant increase), with a typical time lag between the pollutant peak and the event of 2 to 6 days. For other pollution parameters and other types of hospitalizations (e.g., cardiovascular events, eye and skin diseases, etc.), a weaker and ununiform correlations were recorded. Ambient air pollution exposure increases are associated with a short-term increase of hospitalizations due to respiratory tract diseases. The most prominent effect was recorded with the correlation of PM
2.5
and PM
10
. There is only weak evidence indicating that such short-term associations exist between peaks of air pollution concentrations and increased hospitalizations for other (e.g., cardiovascular) diseases. The obtained information could be used to better predict hospitalization patterns and costs for the healthcare system and perhaps trigger additional vigilance on particulate matter pollution in the cities.