The online analysis of volatiles from biomass pyrolysis (or gasification or combustion) is interesting because it has the ability to sample the volatiles directly from their reactive environment. The photon ionization (PI) is an efficient and soft ionization method for online analysis of biomass pyrolysis volatiles. Here, we review recent developments conducted in our groups on PI-mass spectrometry (MS) analysis of biomass pyrolysis volatiles by (1) synchrotron light PI MS and (2) various commercial PI-MS techniques combined with various pyrolysis reactors. The fundamentals of PI MS applied to biomass tar are briefly presented. The effect of photon energy on mass spectra from biomass volatiles is studied by synchrotron PI-MS. Different sources of PI-MS are then compared on vapors produced from fast pyrolysis in a microfluidized bed, namely, argon electron-beam-pumped excimer light (EBEL) vacuum ultraviolet (VUV) lamp single photon ionization (SPI)-MS (126 nm and 9.8 eV), laser Xe cell-SPI-MS (118 nm and 10.5 eV), laser resonance-enhanced multiphoton ionization (REMPI)-MS (266 nm). The suitability of these different ionization techniques for tar online analysis is discussed. The high potential of PI-MS to unravel the mechanisms of biomass pyrolysis is highlighted by some examples of applications. A VUV lamp SPI-MS has been combined to a fixed bed reactor to study the evolution of chemical markers from lignin, cellulose, and hemicelluloses as functions of biomass types and temperature of pyrolysis. It has also been combined to a microfluidized bed to study the fast pyrolysis of different sizes, shapes, and composition of biomass particles. Principal component analysis of the various MS "fingerprints" reveals interesting markers of some pyrolysis regimes.