The scoring system given by Feldman et al. (1966) for their Sexual Orientation Method appears to be based on the unjustified assumption that the scale positions of the adjectival phrases which they use are equally spaced. Consequently their system may treat as inconsistent patterns which in fact are consistent. A class of modifications to the S.O.M. is suggested which remove the need for this assumption, give the same number of consistent response patterns and decrease the number of judgements required of the subject, albeit at the cost of some loss of internal consistency information. © 1968.