This study assessed the possibility of enabling two students with multiple disabilities to request and choose among environmental stimuli through microswitch and computer technology. Each student was provided with two basic microswitches the activation of which made a computer system present a sample of a preferred or non-preferred stimulus. The student could select or reject such a stimulus by vocal utterances discriminated by the computer system as "yes" or "no". If the student selected the stimulus, the computer system presented it for a preset time. If the student rejected the stimulus, the computer system waited for the next input of one of the basic microswitches. Data showed that both students learned to activate the basic microswitches consistently, selected the preferred stimuli about 90% of the times, and largely rejected the non-preferred stimuli. The students also showed some preferences between the two basic microswitches/responses. The performance was maintained at the post-intervention checks carried out 3 and 6 weeks after the end of the intervention. Implications of the findings are discussed technically, practically, and in terms of quality of life.