Julius Caesar (100BC-44BC) is credited as the first person to actually use a cipher in military affairs. His cipher, a shift of the alphabet by three characters, may seem like childs play today, but in those days most people were illiterate and most of their enemies could not read Roman unenciphered. Today, we call a shift of letters by any number a Caesar cipher. Caesar's nephew Augustus, first emperor of Rome, must have thought the cipher was too complex because he used a cipher where each letter was shifted to the very next letter.

A cipher which mixes up the characters in the alphabet is classified as a monoalphabetic cipher. Another example of its use is in the Bible. The Old Testament, in Jeremiah 25:26 and 51:41, uses the name Sheshach in place of Babel. Also, in Jeremiah 51:1 the words "Leb Kamai" or "heart of my enemy" is used for "Kashdim" or "Chaldeans". The cipher is called atbash and substitutes the last character for the first, the second-to-last character for the second, etc. The name atbash is derived from the first 2 and last 2 letters of the Hebrew alphabet since the first two Hebrew letters (aleph and beth) are exchanged for the last two (taw and shin).

The first recorded example of cryptanalysis was many hundreds of years later, by an Arab scholar named Al-Khalil (c. 725-790 AD). He deciphered a Greek cryptogram sent to him by the Bysantine emperor. He explained that he guessed the message would start with "In the name of God" or something similar and worked out the first few characters on that basis, but it took him a month to decipher the message. The words expected in ciphered messages are called cribs, which has been used in cryptanalysis to this day.

It was another 600 years before a more systematic solution to the monoalphabetic cipher was discovered. Ibn ad-Duraihim (1312-1361) was the first person known to use letter frequency analysis. Since the Koran was well-studied and the frequency of each letter was known, he applied this frequency to the frequency of the letters in the ciphered message. He would substitute the most frequent letter in the Koran with the most frequent letter in the ciphered message and then the second most frequent, etc. While doing this, he would see if letter pairs that rarely appear together were being formed, which would mean he would have to pick the next most frequent letter to see if that combination would work.

This is the accepted technique to solve monoalphabetic ciphers today, but this type of cipher is not really in use any more. These ciphers can be solved in mere minutes if the message has enough characters (about 150).