The order of assignment of the Baudot letters to codes is actually alphabetic: first all the vowels plus the accented E (needed, perhaps coincidentally, to write Baudot's first name,) then the consonants, with controls interspersed at the beginning, in a block in the middle, and at the end of the sequence. The ordering of the codes themselves follows a simple reflected-binary Gray code progression, which was actually discovered by Baudot many decades before Gray got a patent on it. Gray codes are well explained elsewhere, e.g. in Wikipedia: http://en.wikipedia.org/wiki/Gray_code#Constructing_an_n-bit_gray_code NOTE: The order of the five-bit Baudot code using Baudot's original bit numbering (54 123) based on the numbering of the keys on the chording keypad effectively reverses the order of the three low-order bits relative to the table given on many websites. Additionally, Gray code does not conform to rectangular tabular layout by normal binary order, although it works fine if you instead lay it out by position in the reflected-binary Gray code sequence (shown below in octal as the "Gray" column.) Baudot Letters Gray Oct 54 123 Letter 00 00 .. ... NUL 01 01 .. @.. A 02 03 .. @@. É (FR), / (UK), or CR (ITA 1) 03 02 .. .@. E 04 06 .. .@@ I 05 07 .. @@@ O 06 05 .. @.@ U 07 04 .. ..@ Y 10 14 .@ ..@ B 11 15 .@ @.@ C 12 17 .@ @@@ D 13 16 .@ .@@ F 14 12 .@ .@. G 15 13 .@ @@. H 16 11 .@ @.. J 17 10 .@ ... FS 20 30 @@ ... ER, often shown as * 21 31 @@ @.. K 22 33 @@ @@. L 23 32 @@ .@. M 24 36 @@ .@@ N 25 37 @@ @@@ P 26 35 @@ @.@ Q 27 34 @@ ..@ R 30 24 @. ..@ S 31 25 @. @.@ T 32 27 @. @@@ V 33 26 @. .@@ W 34 22 @. .@. X 35 23 @. @@. Z 36 21 @. @.. t or c superscript (FR), - (UK), or LF (ITA 1) 37 20 @. ... LS The figures case has a different sequence. Ordering the digits leaves the controls (which are of course unaffected by shifts) and two combinations of the least-significant bits at the ends of each section the sequence. The sequence sections seem to relate to the earlier Hughes code for some details of overall ordering, but ultimately the reason it was chosen is mysterious to me. Strict binary ordering leads to more Hughes-like pairs of adjacent punctuation, but interrupts the sequence of digits with some additional symbols. For reference, the Hughes code: Hughes Figures (× was sometimes replaced with section sign) BLACK 1 2 3 4 5 6 7 8 9 0 . , ; : WHITE ? ! ' + - × / = FS ( ) & " LS Hughes Letters (W was sometimes replaced with É) BLACK A B C D E F G H I J K L M N WHITE O P Q R S T U V FS W X Y Z LS And now the Baudot figures case, in three variations: Baudot Figures (FR) Oct: 0 1 2 4 5 7 3 6 Oct 54 123 ... @.. .@. ..@ @.@ @@@ @@. .@@ 00 .. NUL 1 2 3 4 5 & o 10 .@ FS 6 7 8 9 0 h f 20 @. LS . , ; ! ' : ? 30 @@ * ( ) - / % = No. NOTE: The small o, h, and f are actually superscript, and No. should be the Numero sign. ER is here shown as *, as it often was written. Baudot Figures (UK) Oct: 0 1 2 4 5 7 3 6 Oct 54 123 ... @.. .@. ..@ @.@ @@@ @@. .@@ 00 .. NUL 1 2 3 4 5 1/ 3/ 10 .@ FS 6 7 8 9 0 4/ 5/ 20 @. LS . 9/ 7/ 2/ ' : ? 30 @@ * ( ) - / + = £ NOTE: the fraction numerators are actually superscript. ER is here shown as *, as it often was written. Baudot Figures (ITA 1) Oct: 0 1 2 4 5 7 3 6 Oct 54 123 ... @.. .@. ..@ @.@ @@@ @@. .@@ 00 .. NUL 1 2 3 4 5 CR (1) 10 .@ FS 6 7 8 9 0 + (2) 20 @. LS LF , . (3) ' : ? 30 @@ * ( ) - / % = (4) NOTE: ER is here shown as *, as it often was written. The four national-use code positions are shown as (1), (2), (3) and (4). I found a web page with an old scanned table showing the "proper" figures ordering. There is a method to the grouping: binary order for the high-order bits 54, next grouping by number of set low-order bits 123, then grouping by each low-order bit pattern (irrespective of shift; sorted in binary order) and finally sorted by shift. The website: NADCOMM Papers and Writings: Five-unit codes by Alan G Hobbs, G8GOJ President of British Amateur Radio Teledata Group, (c) 1999 http://www.nadcomm.com/fiveunit/fiveunits.htm It links to the following scan of an old Baudot chart, showing both the French and British variations, the former in alphabetic order (although the vowels have been sorted in among the consonants, so the Gray code progression is obscured,) and the latter in the "natural" figure order, which matches the earlier Hughes code very closely. The scan: http://www.nadcomm.com/fiveunit/baudot.gif And here is the "natural" order for figures (1 2 4 3 6 5 7 0 in octal): Baudot Figures (FR) Oct: 1 2 4 3 6 5 7 0 Oct 54 123 @.. .@. ..@ @@. .@@ @.@ @@@ ... 00 .. 1 2 3 & o 4 5 NUL 10 .@ 6 7 8 h f 9 0 FS 20 @. . , ; : ? ! ' LS 30 @@ ( ) - = No. / % * For most of the figures present in the Hughes code, the order in Baudot's code is identical.