This study aimed at examine the number of planned and acute hospital admissions during 1 year among people 65+ and its relation to municipal care, outpatient care, multimorbidity, age and sex. Four thousand nine hundred and seven individuals having one or more admissions during 2001 were studied. Data were collected from two registers and comparisons were made between those having one, two and three or more hospital stays and between those with and without municipal care and services. Linear regression was used to examine factors predicting number of acute and planned admissions. Fifteen percent of the sample had three or more hospital stays (range 3-15) accounting for 35% of all admissions. This group had significantly more contacts in outpatient care with physician (median number of contacts (md)=15), compared to those with one (md: 8), or two admissions (md: 11). Main predictors for number of admissions were number of diagnosis groups and number of contacts with physician in outpatient care. Those who are frequently admitted to hospital constitute a small group that consume a great deal of inpatient care and also tend to have frequent contacts in outpatient care. Thus interventions focusing on frequent admissions are needed, and this requires collaboration between outpatient and hospital care.