Although formulaic expressions found in earlier correspondence have drawn scholarly attention, their (un)grammaticality has not been thoroughly researched. The present paper thus focuses on the two types of formulae with the verb remain found in private correspondence: one headed by 1st person pronoun (as in: we remain(s) your daughters), the other one starting with but/so/also/and/only (as in: but remain(s) your affectionate child until death). For the purpose of the study a corpus of 19th-century correspondence has been compiled and analyzed; additionally, the data from Dylewski (2013) have been taken into account. Next to the corpus scrutiny, an Internet search has been carried out to verify whether the use of the formulae at issue goes beyond the 19th century. An analysis from both a qualitative and quantitative angles allowed for putting forth a number of hypotheses concerning the origin of variation between -s-marked and unmarked forms as well as their distribution across letter-types and different geographical locations. The results of the analysis also corroborate the claim that -s on remain in the structures under discussion is neither a “part of the authentic local vernacular nor of authentic contemporary standard English, but part of a specifi c, localized practice of letter writing, which had its own linguistic rules” (Pietsch 2015: 226).