This study examines the use of orthographically reduced forms in messages posted on the microblogging site Twitter (tweets). A total of 900 tweets all related to current event discussions were collected and analysed. By comparing data from tweets subject to a maximum limit of 140 characters with tweets posted after a doubling in the character allowance, the study aims to explore to what extent character restrictions influence the use of reductions. Symbols and clippings were found to be the most frequently used types of reductions, with the ampersand symbol appearing to be a standard feature of tweets. The overall rate of non-standard reductions for the entire sample was 2.14% of all words, with the sample subject to a lower character allowance showing a markedly higher reduction rate compared to the higher character allowance sample. This supports the idea that character constraints affect users’ choice between standard forms and non-standard reduced spellings, a finding which may also be applicable to SMS messages. The very low rate of non-standard reductions may also indicate a preference towards the use of more formal language among Twitter users.