The ease of adjustability of 'ergonomic chairs' was investigated. Twenty-six chairs were classified with respect to control features. There were 24 different types of control arrangements. Twenty subjects participated in a series of three experiments. The first experiment investigated what types of control mechanisms could be identified by naive users who had received no training or instruction in the use of control devices. The chairs were classified into five broad categories, and one chair from each category was tested. Two control design features, control discernibility and control feedback, significantly improved users' understanding of the adjustability controls. In the second experiment controls on 20 chairs were evaluated. Subjects were first trained in the operation of the controls for each chair; and then they performed a chair adjustability task. Six control evaluation criteria were used: reachability, force, ease of operation, preference, time to adjust, and number of control activations. Controls with long levers were preferred to controls with short levers or push buttons. The third experiment investigated the effect of the number of adjustability features on ratings of comfort and time for adjustment. Four chairs with different levels of control complexity were evaluated. The chair with the greatest number of adjustability controls took significantly greater time to adjust and was also judged to be the most comfortable.