Objective To study the level of carer reported distress in mild dementia, over a 3‐year period. Methods This study is part of the Norwegian DemVest‐study and utilises data from carers of people with mild dementia (n = 223). Those diagnosed with dementia with Lewy bodies (DLB, n = 63) and Alzheimer's disease (AD, n = 97) were included together with other dementia types (n = 63). The Relatives' Stress Scale was used to assess the level of reported distress in carers. Descriptive and a linear mixed effects models including diagnosis, time, and the interaction between time and diagnosis were performed. Results Carer distress in mild dementia increased significantly over time (P = 0.011), particularly from baseline until 2 (P = 0.001) years follow‐up. Carer distress in people caring for those with AD increased significantly, from baseline until 2 (P = 0.047) and 3 (P = 0.019) years follow‐up. Distress in carers of people with DLB was high at baseline and remained relatively stable across the 3‐year period. However, admission to a nursing home during the first year of follow‐up was associated with a significantly lower reported carer distress in those caring for a person with DLB (P = 0.002), compared with those caring for a person with DLB living at home. Conclusion Being a carer to a person with mild dementia is associated with increasing distress. However, the burden of distress changes with the diagnosis, time, and situation, which highlights the dynamic nature of the caring role. Findings have important implications for health services for people diagnosed with mild dementia and their carers.