Forty-eight subjects (24 experienced and 24 inexperienced) entered name and address data on three types of keyboard layouts, the standard QWERTY, an alphabetic, and a random layout (as a control condition). Each keyboard was used in a squared arrangement where the keys were arranged in perfectly aligned rows and columns, and an offset version, where for every other row, the columns were slightly offset. The data show both preference and performance advantages for the QWERTY arrangement. Recommendations for design are discussed