ABSTRACT Thermogravimetry (TG) is used to measure the change of sample mass versus temperature or time under a program controlling temperature. It is a common thermal analysis tool with a precise built-in balance, and accurate data can be obtained through a small amount of sample at the level of milligrams. The feasibility of the TG method for determining maximum water content (MWC) measurement of waterlogged archaeological wood was studied. Simulated waterlogged wood and waterlogged archaeological wood were used to determine the test condition for the TG method. MWC data obtained through the oven-dry and TG methods with different mass gradients from 0–10 mg to >50 mg were analyzed statistically. Deviation, error bar, and coefficient of variation were evaluated. According to the results, when sample mass is no less than 30 mg, and holding time at 105°C is no less than 40 min, data deviation between these two methods is less than 5%. Effect of sampling depth and presence of inorganic deposits were also investigated. Specimens were taken from an archaeological wood with a sampling depth from 2 to 6 mm, then MWC was measured through the TG method. The results indicate MWC of the samples taken from a depth of 2–4 mm is higher than that taken from a depth of 0–2 mm and is lower than that taken from a depth of 4–6 mm, which is related to inorganic deposits. The presence of deposits was verified through SEM-EDX. MWC of three archaeological wood samples from China and five archaeological wood samples from Italy were measured to verify the effectiveness of the TG method when wood species and degradation degree are different.