We analysed the data on 876 patients after lower limb amputation who were admitted to inpatient rehabilitation at the University Rehabilitation Institute in Ljubljana from 2015 to 2017. Upon admission, they were asked about falls at home and/or in the acute hospital since the amputation. The data on comorbidities was also recorded. Six months after discharge, the patients were asked about falls in their home environment. In addition, a test group took part in a short education during inpatient rehabilitation about fall prevention. We applied tobit, negative binomial, logistic, Poisson, zero-inflated Poisson and proportional odds regression models to assess the effect of potential fall risk factors. All regression models were fitted with main effects only. The models for falls during inpatient rehabilitation were adjusted for the length of hospital stay (either as offset variable or as predictor). Different regression models with different assumptions led to similar conclusions. A short-term effect of the education program was observed. Unilateral amputation, age, female sex, falls before admission to inpatient rehabilitation and stroke and/or myocardial infarction proved to be the main risk factors for falls at various stages after lower limb amputation. [ABSTRACT FROM AUTHOR]