An analysis of the careers of serial rapists indicates that for most there was no significant change in the amount of force used, the pleasure experienced by the rapist, the extent of victim injuries, or the duration of the assault over the first, middle, and last offenses. This observation is consistent with the hypothesis that the rapist's behavior during the assault is determined more by the rapist's modus operandi than the varying interaction between the perpetrator and his victim. There is, however, a dangerous minority who escalate their use of violence over successive offenses. This group of rapists is unusually prolific. Victim resistance was found not to vary with the amount of force used by the perpetrator, although it was associated with the pleasure experienced by the rapist and the duration of the rape. The classificatory analyses demonstrate that behavioral scales can be used to classify rapists both according to rape type and "increaser" status. 3 tables and 27 references