Wood-based materials, which are more and more used in building construction, are often characterized by the normative tests such as cone calorimeter. Measurements of ignition time, mass loss or heat release rate, and temperatures are usually performed. The degradation dynamics of samples depends on a large number of variables (humidity, density, scale sample, experimental conditions), and it can induce flaming ignition. The autoignition phenomenon has already been studied, but opinions concerning the involved physical processes remain controversial. The ability of wood to self-extinguish in case of fire is also an important issue, especially concerning structure mechanical strength. The purpose of the present work is to study thermal degradation of glulam in the cone calorimeter with the dedicated experimental setup. In particular, we focus on ignition, extinction, and re-ignition. The cone calorimeter with vertical sample orientation was modified to study extinction and re-ignition so that the external heat flux can be quickly switched off. The samples were exposed to the heat fluxes at four values in the range from 15 to 75 kW/m2. For the low heat flux, the transition was observed between pyrolysis and smoldering combustion at the temperature about 400 °C. The ignition only occurred at the heat flux higher than 55 kW/m2. For 55 kW/m2, re-ignition was not consistent and random. However, for the highest flux, reignition occurred for each new exposure to the external heat flux. Both the ignition and extinction times increased at each exposure. Char layer delays the ignition.