The effects of supplementary computer instruction in house dust mite-avoidance measures on adherence to implementing measures, on home dust mite-allergen levels, and on symptomatology were investigated in 52 adult patients with mite-associated asthma. Twenty-six patients received conventional instruction (counseling and written instruction) and the other 26 patients received conventional plus 22 minutes of interactive computer-assisted instruction. Instructions were aimed at mite-avoidance measures. Pre- and postinstruction dust samples were collected, and adherence was monitored. All patients kept symptom diaries twice a day. Patients' progress was followed for 12 weeks, and all patients completed the study. Adherence, number of observed and self-reported mite-avoidance measures implemented after visit, was higher for the computer group (p = 0.023). The computer-instructed group achieved significantly lower levels of mite allergen in bedroom carpets (p = 0.004) with mean levels of mite allergen declining from 6.5 +/- 7.6 to 2.2 +/- 4.3 micrograms/gm of dust (two-site monoclonal antibody assays), whereas levels for the conventional-instructed group did not change. Moreover, by study weeks 9 and 10, the computer-instructed group was significantly less symptomatic (p = 0.033). Mean symptom scores for this group decreased from 12.4 to 7.7, compared with 16.4 to 14.3. Conventional instruction supplemented with computer instruction is suggested in mite education.