Flash memory is a non-volatile computer storage device which consists of blocks of cells. While increasing the voltage level of a single cell is fast and simple, reducing the level of a cell requires the erasing of the entire block containing the cell. Since block-erasures are costly, traditional flash coding schemes have been developed to maximize the number of writes before a block-erasure is needed. A novel coding scheme based on error-correcting codes allows the cell levels to increase as evenly as possible and as a result, increases the number of writes before a block-erasure. The scheme is based on the premise that cells whose levels are higher than others need not be increased. This introduces errors in the recorded data which can be corrected by the error-correcting code provided that the number of erroneous cells are within the error-correcting capability of the code. The scheme is also capable of combating noise, causing additional errors and erasures, in flash memories in order to enhance data reliability. For added flexibility, the scheme can be combined with other flash codes to yield concatenated schemes of high rates.