Three epoxy systems (DGEBA + mPDA, TGDDM + DDS, and Fiberite 934(TM)) were used to investigate glass transition temperature (T-g) variation of epoxy under hygrothermal environment exposure. Materials were immersed in distilled water at constant temperatures of 45 degrees C, 60 degrees C, 75 degrees C, and 90 degrees C for water absorption and then desorbed at different temperatures. Thermomechanical analysis (TMA) and differential scanning calorimetry (DSC) were employed to determine T-g changes at different hygrothermal stages. The investigations revealed the following results: i) the change of T-g does not depend solely on the water content absorbed in epoxy resins, ii) T-g depends on the hygrothermal history of the materials, iii) for a given epoxy system, higher values of T-g resulted for longer immersion time and higher exposure temperature, and iv) the water/resin interaction characteristics (Type I and Type II bound water) have quite different influence on T-g variation. A sorption model and collateral evidence introduced in Part I of the series were used to interpret and explain T-g variation in epoxy resin systems. Both Type I and Type II bound water influence T-g variations, albeit in different ways. Type I bound water disrupts the initial interchain Van der Waals force and hydrogen bonds resulting in increased chain segment mobility. So Type I bound water acts as a plasticizer and decreases T-g. In contrast, Type II bound water contributes, comparatively, to an increase in T-g in water saturated epoxy resin by forming a secondary crosslink network. The experimental T-g values encompass the combined effect of the two water-resin interaction mechanisms described briefly in the preceding text and in detail in Part I of this paper series. The often-cited polymer-diluent model used to predict T-g variation of polymers exposed diluent media is lacking when a dual sorption mechanism is involved during hygrothermal exposure process.