To evaluate the compression wood response in eastern white pine (Pinus strobus L.) following a severe ice storm in 1994, 47 trees were felled in 2007 and cross-sectional samples were collected at 0.5 (±0.2) m stem height. The disks were sanded and digitally scanned, and the cross-sectional area (mm2) of compression wood within each tree-ring was quantified using image analysis software. Topographic data (slope, aspect, and elevation) were also recorded for each P. strobus tree, along with a modified competition index. Wood anatomical features were also quantified in the three years before and after the storm along a tree diameter gradient. Although tree age was relatively constant in this stand, tree size was influenced by topographic position; larger trees grew in the valley while smaller trees were found growing in thin soils at the mid-slope position. When the cohort was about 25 years old, ice deposition caused a heterogeneous compression wood response which was highly related to tree size. In the thirteen years following the ice storm, the 6 – 9 cm (2007) diameter class formed significantly more compression wood area than any other, followed by the 10 – 13 cm (2007) diameter class. The tree diameter range that formed the most post-storm compression wood was 4 – 8 cm at the time of the storm, suggesting that this diameter range was most affected by 8.5 cm of ice loading in P. strobus. Trees > 18 cm in 1994 did not form any compression wood after the storm, but many experienced a growth release to fill canopy gaps. Topographic variables did not influence compression wood formation directly, but only one plot was sampled so these results are tenuous. However, topography did influence tree size which was the most important predictor in compression wood. There was no relationship between compression wood area and competition index. Due to compression wood formation after the ice storm, cell wall thickness and cell circularity were significantly higher in the 1994 tree-ring than in other rings examined (1991 – 1993, 1995, and 1996). Tracheid and lumen diameters were significantly smaller in compression wood cells (30.5 and 19.5 μm, respectively) than in normal wood (36.8 and 28.4 μm, respectively); opposite wood cells were intermediate in size (32.4 and 24.4 μm, respectively). Due to small tracheid size, compression wood contained significantly more cells mm⁻¹ (33) than normal wood (27), but no significant differences in cell wall area. Therefore, cumulative cell wall area occupied 47% of the cross-section in compression wood tissue on average, compared to 31% in normal wood. Dispersing tree weight across a greater surface area may help compression wood to prop up a bent tree, but reduced lumen area may also impact hydraulic conductivity in the stem. / Master of Science