OBJECTIVES: Caregiver burden is a key measure in caregiver research and is frequently used as a baseline measure in intervention studies. Previous research has found numerous factors associated with caregiver burden such as the relationship quality between carer and patient, the patient's cognitive ability, behavioural and psychological symptoms displayed by the patient, caregiver gender, adverse life events to name a few. Many studies have investigated these factors singularly however current thought suggests a multi-factorial role and inter-dependence of these factors. Based on this it was decided to investigate factors associated with caregiver burden using a multiple regression analysis in order to ascertain the predictive quality of these factors of caregiver burden. METHOD: Cross-sectional study using validated measures of a patient's cognitive ability, ability to carry out day-to-day tasks and behavioural and psychological symptoms. Caregiver outcomes used are caregiver burden, relationship quality, caregiver confidence, experience of adverse life events, neuroticism, age and gender. Interviews and questionnaires were carried out on 74 patients diagnosed with dementia and their main caregivers from the Midlands of England. RESULTS: Multiple regression analysis showed that caregiver overload, carer-patient relationship quality, the experience of adverse life events, caregiver gender, caregivers' level of neuroticism, caregiver role captivity and the level of caregiver confidence accounted for over 80% of the variance in caregiver burden. CONCLUSION: These results confirm previous correlational research on caregiver burden. Furthermore, due to the use of multiple regression analysis the findings also show factors that are clear predictors of caregiver burden and we offer possible suggestions from these findings on future clinical practice interventions on caregiver burden.