Objective Although disabled elderly people mostly prefer to receive care at home or in other community settings, many of them reside in nursing homes. That is why several researchers have tried to identify predictors of institutionalisation. Various different dependency factors seem to explain the request for residential care. The aim of this study is to discover the most important factor triggering an urgent request for nursing home admission. Methods On the basis of social field research, we analysed the profiles and motives of an admission cohort of 125 elderly (31 men and 94 women) who were admitted to four nursing homes in Antwerp (Belgium) between January 2000 and April 2001. The study used data of the 'intake conversation', performed by an experienced social worker of the nursing home, subsequent to the request for nursing home admission. Gender, age, Katz category, marital status, disease, living conditions, Personal and Instrumental Activities of Daily Living (P-ADL and I-ADL) were the independent variables. The variable 'time span' was introduced as dependent variable. This is the time between the onset of dependency and the request for institutionalisation. Nursing home carers have classified this time span in three periods: < 3, 3-12, and > 12 months. The statistical analysis focused on the characteristics of the two extremes, namely the earliest versus the latest applicants (n = 74). This was the best strategy to go about investigating the issue due to the vagueness and uncertain status of the data in the midrange. Results The applicants had an average age of 83 years. 31% of the elderly were defined as functioning good (needing assistance from another person in no to maximum two ADLs - washing and dressing) and 69% were classified as ill functioning (needing assistance in minimum three ADLs). Women were more likely to be widowed (83%) and to live alone and isolated (55%) and they had a lower degree of dependency (both P-ADL and I-ADL) when entering institutions. Of the women, 57% had a mental illness, compared with 48% of the men. Of the applicants, 34% were unwilling or unable to start home care and applied for an urgent request (within the first 3 months); 41% tried home care for a time and 26% applied after one year of home care. The stepwise logistic regression analysis identified I-ADL as the decisive factor explaining the difference in 'decision speed' towards institutionalisation. An increase of one unit on the I-ADL score increased the chance of a request within the first three months by 63% (95%CI: 19 to135%, p = 0.006). Conclusions The only factor related to an urgent request for nursing home admission seems to be the I-ADL score. These results have important implications for targeting sheltered housing and further extension of home care services to postpone or prevent institutionalisation.