Two parallel experiments investigated transfer to instrumental eyelid conditioning following 0, 5, 10, or 20 classical reinforcements with the same CS and transfer to classical eyelid conditioning following 0, 5, 10, or 20 instrumental reinforcements that were contingent on .Ss emitting an eyelid response that resembled a classically conditioned CR upon the appearance of the CS.Amount of transfer to instrumental conditioning was a progressively increasing function of the number of pretransfer classical reinforcements, and the instrumental acquisition curves were similar to those previously obtained. Also, the conditioned (C) or voluntary (V) classification of the instrumental response (IR) form resembled the response form that was recorded during pretransfer training. Transfer from instrumental to classical conditioning was not influenced by the number of pretransfer instrumental reinforcements, the classical acquisition curves were abnormal, and the response form in the transfer phase was not related to the response form recorded in the pretransfer phase. It is concluded that transfer from classical to instrumental training occurs readily and is a function of the number of pretraining classical reinforcements, but that a parallel transfer from instrumental to classical conditioning may be more difficult to demonstrate if it does occur.