Despite the increasing outdoor use of fire-retardant-treated wood, methods for predicting its service life remain poorly established. With the aim of establishing a method to predict chemical losses from fire-retardant-treated (FRT) wood caused by humid atmospheres and rain outdoors, this study examined the preferable conditions for artificial-weathering tests and demonstrated the acceleration coefficients in these tests (i.e., the ratio of equivalent time to reach the same retention of chemicals in artificial weathering and outdoors) based on the EN927-6 standard. To determine the moisture absorption and desorption levels of FRT exposed to outdoor conditions, an outdoor exposure experiment was conducted. The moisture content was higher in the FRT wood than in untreated wood, regardless of the type of coating, and ranged between 11% (in March) and 50% (in September) of the untreated wood’s weight. EN927-6 artificial weathering tests were performed on two groups of specimens with initial moisture contents of 0% and 25%. Retention rates of fire-retardant chemicals after a 2520-h test were compared with those retrieved from 4-year outdoor exposure reported elsewhere. Comparison of these two experiments demonstrated that the acceleration coefficients were 4.1–11.3 in the case of specimens with 0% initial moisture content and 5.1–11.4 in the case of specimens with 25% initial moisture content. The higher initial moisture content produced a more uniform acceleration coefficient. Nevertheless, larger acceleration coefficients were derived from specimens with penetrating or semi-film-forming coatings in both cases. The relationships between the uniformity of this acceleration coefficient and the initial moisture content are discussed from the moisture absorption experiment under constant temperature and humidity and under condensation conditions.