Many automated data-handling systems still require the handprinting of entries on special forms as an initial step. This investigation sought and evaluated methods for bypassing handprinting in the manual entry of data into computers. Three laboratory experiments were performed to obtain data on human performance rates in various input modes, including writing, printing, marking, and keying with both print and scope feedback. It is estimated that conversion to new input methods might initially slow down the input rate of analysts who formerly handprinted their entries but that practice would be likely to restore former speeds.