/// <summary>
/// Convert the argument string into its binary-coded decimal (BCD) representation, e.g.
///  "0110" -> { 0x01, 0x10 } (for Big Endian byte order)
///  "0110" -> { 0x10, 0x01 } (for Little Endian byte order)
///  (NOTE: It is assumed that the string is always a big-endian representation.)
/// </summary>
/// <param name="isLittleEndian">True if the byte order is "little end first (leftmost)".</param>
/// <param name="bcdString">String representation of BCD bytes.</param>
/// <returns>Byte array representation of the string as BCD.</returns>
/// <exception cref="ArgumentException">Thrown if the argument string isn't entirely made up of BCD pairs.</exception>
public static byte[] ConvertToBinaryCodedDecimal(bool isLittleEndian, string bcdString)
{
    bool isValid = true;
    isValid = isValid && !String.IsNullOrEmpty(bcdString);
    // Check that the string is made up of sets of two numbers (e.g. "01" or "3456")
    isValid = isValid && Regex.IsMatch(bcdString, "^([0-9]{2})+$");
    byte[] bytes;
    if (isValid)
    {
        char[] chars = bcdString.ToCharArray();
        int len = chars.Length / 2;
        bytes = new byte[len];
        if (isLittleEndian)
        {
            for (int i = 0; i < len; i++)
            {
                byte highNibble = byte.Parse(chars[2 * (len - 1) - 2 * i].ToString());
                byte lowNibble = byte.Parse(chars[2 * (len - 1) - 2 * i + 1].ToString());
                bytes[i] = (byte)((byte)(highNibble << 4) | lowNibble);
            }
        }
        else
        {
            for (int i = 0; i < len; i++)
            {
                byte highNibble = byte.Parse(chars[2 * i].ToString());
                byte lowNibble = byte.Parse(chars[2 * i + 1].ToString());
                bytes[i] = (byte)((byte)(highNibble << 4) | lowNibble);
            }
        }
    }
    else
    {
        throw new ArgumentException(String.Format(
            "Input string ({0}) was invalid.", bcdString));
    }
    return bytes;
}

BCD is where the hex value of a byte represents two integer values between zero and nine; e.g., 0x98 -> "98" (whereas its integer value is actually 152). The method throws an exception if non-BCD data is input.

You can also check out the partner method, [Convert bytes into a binary coded decimal (BCD) string](http://snipplr.com/view/16802/convert-bytes-into-a-binary-coded-decimal-bcd-string/).

Update:    
Unfortunately, there doesn't seem to be a built-in method for doing this (there is for the converse; see System.BitConverter.ToString). The System.BitConverter.GetBytes(char) overload returns the byte value of the character, rather than a BCD byte array (`BitConverter.GetBytes('A')` returns a byte array holding `[65, 0]`, not [0x0A] as we would want).

Convert a binary coded decimal (BCD) string into bytes

C#