OBJECTIVE： To provide reference for better carrying out safe medication intervention for the empty-nest elderly. METHODS： Two communities with different levels of income and culture were collected from Pudong New District of Shanghai， i.e. new community A and traditional community B. By cluster sampling， 69 and 110 empty-nest elderly were selected as respondents respectively. The questionnaire survey about drug storage and medication habits of the elderly was conducted. The data of the survey were analyzed statistically. RESULTS： A total of 69 and 110 questionnaires were sent out， and 44 and 63 questionnaires were recovered with effective recovery rate of 63.8％ and 57.3％， respectively. The rate of drug storage in families from traditional community B was higher than new community A （100％ vs. 88.6％，P＝0.010）. The proportions of residents in two communities storing their drugs in special drawers were the highest （59.0％ and 57.1％）. Nearly half of the surveyed households in 2 communities did not form the habit of regularly clearing stored drugs. The frequency of residents who cleared stored drugs in traditional community B was higher than those in new community A （P＝0.009）. Nearly more than 70％ of the residents threw their expired drugs to garbage and only 11％ to drug stores or community health service center. The proportion of residents who were used to reading drug instructions in new community A was higher than traditional community B （95.5％ vs. 77.8％，P＝0.024）. The proportion of the residents consulted community doctors when they were puzzled by drug use in new community A was lower than traditional community B （56.8％ vs. 76.2％）； The proportion of the residents consulted physicians in major hospitals was higher than traditional community B （34.1％ vs. 9.5％）； there was statistical significance （P＝0.034）. The most popular health service for the residents in the new community A was the hotline for the drug use （65.9％）， while the residents in traditional community B needed blood pressure and sugar monitoring service and set up the health records （41.3％）. The proportion of the residents’ following demands in new community A was higher than traditional community B： opening drug use hotline， publicizing the knowledge of diseases and medication and conducting lectures （P＜0.001，P＝0.006，P＝0.018）. CONCLUSIONS： The proportion of drug storage in empty-nest families of two communities is in very high level， but drug storage methods， cleaning frequency are not satisfactory. There are still a small number of elderly people who do not have the habit of reading drug instructions. There are some differences between the old and new communities. At the same time， the needs of health services in new and old communities are different.