The work presents the results of short-term health effects assessment of particulate matter (PM) in Warsaw, the capital of Poland. The influence of three PM fractions, PM10 (particles of aerodynamic diameter < 10 μm), PM2.5 (particles of aerodynamic diameter < 2.5 μm), and PMc (coarse fraction of diameter between 2.5 and 10 μm), modeled by the CALMET/CALPUFF system, has been studied in the period of 2013–2014. Six population health endpoints; daily counts of all-cause (ALL), cardiovascular (CV), and respiratory (RS) death cases; and ALL, CV, and RS hospital admissions were investigated with the use of statistical time series analysis via nonparametric generalized additive model (GAM) approach. The results show that PM2.5 increases the relative risk (RR) of ALL premature deaths by 0.7% per 10 μg/m3, as well as of CV mortality by 0.9%. PM10 exposures reveal the largest influence on mortality in a 2-day lag: 0.3% for all causes and 0.4% for CV causes, while for RS causes only in the elderly group (above 65 years, 1.4%) and for males (2.1%). The risk of hospitalizations increases with elevated PMc levels by 2.5%, 2.1%, and 4.6% for ALL, CV, and RS hospital admissions, respectively. The results suggest that the research on PM impact on health should concentrate more on attempts to assign specific health outcomes to PM originating from different types of sources, characterized by different granulation, as well as physical and chemical properties of emitted particles.