Recognizing that the transfer of bedridden patients is associated with a high rate of low back injuries, various devices have been developed to assist with sparing the patient handlers. The purpose of this study was to quantify the friction-reducing ability of three different 'sliding' patient transfer devices together with the subsequent consequences on the low back loads of people performing the transfers. Coefficients of friction of the devices were determined by 'transferring' a standard object and a 'patient' over several surfaces common to a hospital setting. Then three participants performed controlled transfers with the various devices. Electromyography to measure muscle activation levels together with external forces and kinematic positional data were collected during push, pull and twist transfers. Spine loads were estimated with a three-dimensional biomechanical static link-segment model of the human body. Simply sliding a patient on a cotton sheet (control condition) produced a coefficient of friction of 0.45. The assistive devices substantially reduced friction by well over one-half (coefficients of 0.18 - 0.21). However, when using the devices the subjects adopted a variety of postures and techniques, such that there were no consistent influences on trunk inclination, low back compression or muscle activation profiles. Direct measurement of reduced friction between the bed and the patient with a friction-reducing device together with measurement of the back loads when actually transferring a patient formed a proof of principle. Specifically, while the device lowers friction, the transfer technique adopted by the lifter must be proper to reduce low back loading and any subsequent risks of back troubles associated with patient transfers. The direction of hand forces and torso position remains important.