A 3-2 pulldown detector for a video compression encoder initially converts each frame of an input video signal that includes converted film material as well as original video material into top and bottom fields. Consecutive fields from the top and bottom fields are then input to respective sum of absolute difference circuits, the outputs of which are subtracted from each other to produce a difference signal. The difference signal is processed to remove spikes that represent scene changes and to generate a flag signal during the periods of the input video signal when the converted film material is present. The flag signal is then used by a frame rate reduction circuit to eliminate the repeated fields in the converted film material prior to input to a video encoder that outputs a compressed video signal.