To explore the risks of long-term proton pump inhibitors (PPI) use. A total of 79 study subjects were selected from patients seeking health care at outpatient gastroenterology clinic at Peking University People's Hospital during January 2011 to June 2013. Long-term PPI use was defined as taking PPI for at least one year. They were divided into study (n = 79) and control (n = 47) groups after matching for age and gender. The study group had 43 females and 36 males with an average age of (63 ± 10) years. And the control group had 28 females and 19 males with an average age of (63 ± 9) years. Data collected were demographic characteristics, clinical diagnosis, detailed dosing information of PPI. serum hemochrome, serum gastrin, vitamin B(12), serum iron, magnesium, calcium, copper, zinc, bone density of hip joint and lumbar vertebra L(1-4). Statistic analysis was performed with SPSS software (t and χ(2) test). The medium duration of PPI use was 3.07(1-15) years. Daily usage was a full or half amount of pharmaceutical directed dosage. Questionnaires regarding appetite, defecation, infection, otalgia change in long-term PPI use indicated:among them, 77(97.5%) cases had no appetite change and 2 (2.5%) improved. And 73 cases (92.4%) had no change or 2 cases improved (2.5%) in bowel movements habits while another 4 cases (5.1%) changed into loose stools. Lung or intestinal infection was not found. No case had an onset of otalgia related with PPI intake. Among PPI users, 14 cases (17.7%) appeared to have gastric polyps. T value of total bone density of hip and inter-hip was -0.93 ± 1.04 and -0.61 ± 0.91 respectively in the study group. And both were lower than those in the control group (-0.44 ± 1.12, -0.19 ± 0.94; t = -2.49, -2.35; P = 0.014, 0.021). No statistical differences existed in T value of neck of hip, as well as total T value and T value of each L(1-4) between study and control groups (all P > 0.05). Higher serum Mg compared to control group ((1.03 ± 0.08) vs (0.97 ± 0.07) mmol/L, t = 4.13, P = 0.000). No difference of serum calcium, copper, iron, zinc, vitamin B(12), folic acid and Hb,MCV,MCH between study and control groups (all P > 0.05). Higher serum gastrin versus normal range and controls were found in the study group ((108 ± 28) vs 100 vs (79 ± 26) ng/L, t = 2.56, 5.78; both P < 0.05). Taking manual routine dosage have higher hip and lumbar osteoporosis than half-dosage, and bone density decreased and normal bone density ratios are lower(all P < 0.05). Long-term use of PPI with small or average dosage may slightly decrease hip density and increase serum gastrin. However, there is little effect on vitamin B(12) and minerals.