It is shown that the orthorhombic unit cell dimensions of a given polyethylene are not unique, but depend on the physical history of a sample. For polymer crystallized from a ¾% p‐xylene solution, the values of a and b measured at 23°C by powder camera technique vary up to 0.8% for crystallization temperatures of 50°, 70°, and 90°C, annealing temperatures from 50° to 127°C, and deformation. The as‐crystallized values of a and b as well as the effects of annealing and deformation are smaller the thicker the crystals. For ``undeformed'' crystals the changes correlate with the reciprocal of crystal thickness but for deformed crystals the changes occur before the thickness increases (even below the crystallization temperature). The dimensions undergo the greatest change within the first 2 min of annealing and change very little with time thereafter. The only exception observed is a slow decrease of b with time of annealing at 100°C. Crystals grown from other solvents exhibit very nearly identical effects, and the presence of solvents does not alter the cell dimensions at room temperature. Similar effects are observed when crystals grown from the melt by slow cooling, quenching in ice water, and quenching in a dry ice‐acetone suspension are annealed at temperatures from 60° to 127°C. In these crystals also the dimensions correlate with the reciprocal of the long period and extrapolate to values of 7.372 and 4.933 Å for a and b, respectively, at infinite long period. The origins of the effects cannot be assigned with certainty but changes in the fold surfaces and their concentration probably play a role; a discussion relevant to this and other mechanisms is presented. Corresponding to the maximum changes in cell dimension are changes of 1% in cell density which corresponds to an apparent change in degree of crystallinity of 6%. These changes are too small to account for the frequently reported discrepancy between the cell density and that measured by pycnometry, etc. However, the changes can account for some of the variation with crystallization temperature of the density measured by pycnometry, etc. The effects of cell variation should be considered in the detailed interpretation of the variation of physical properties with physical history of polymer crystals.