Soft single photon ionization-time-of-flight mass spectrometry (SPI-TOFMS) and principal component analysis (PCA) were applied for the characterization and discrimination of the chemical patterns of all individual cigarette puffs from the 2R4F University of Kentucky research reference cigarette. The SPI-TOFMS was connected to a smoking machine, and 10 cigarettes were smoked under defined smoking conditions. A total of 41 detected mass signals could be clearly assigned to smoke constituents (e.g., unsaturated hydrocarbons, aromatic species, sulfurous compounds, and nitrogen-containing substances). For further analysis, the on-line recorded mass signals were added up for each cigarette puff resulting in a single summed mass spectrum for each puff. The so-achieved puff-by-puff resolved yields were additionally normalized by the corresponding total ion signal, which eliminated the influences of varying amounts of burnt tobacco. These values were incorporated into a PCA to find differences and similarities in the chemical patterns of the individual cigarette puffs. In addition, absolute (without normalization) and normalized puff resolved yields were used to clarify occurring trends. Thereby, it was shown that the chemical pattern of the first cigarette puff was very unique, whereby extraordinary high yields of unsaturated hydrocarbons are mainly responsible for this. Depending on the smoking procedure, the chemical pattern of the second puff can also be separated from the first and the third puff. In this case, nitrogen-containing substances play an important role. Puffs three to eight show only small but observable differences. These changes are greater influenced by oxygen-containing and sulfurous smoke constituents. The findings reveal that the overall chemical patterns of machine-smoked cigarette puffs vary quite a lot during the smoking process. This lets us assume that the burden of hazardous compounds for the human smoker also differs from puff to puff.