With the population aging in many countries, the travel habits of older people are receiving more attention in the transportation literature. However, our understanding of factors in uencing their mode choices is still limited. is research focuses on mode choice for shopping trips, as these are the most frequent trips of older people. e study is not limited to trips to retail locations, but investigates the combined mode choice of trips to and from shops in order to understand which factors in uence mode changes. Two types of models, multinomial logit (MNL) and nested logit (NL), are tted to data from the London Area Travel Survey. e nesting structure is used to test the correlation in mode choice before and a er shopping. A particular focus of the models is on the importance of accessibility variables such as bus and rail stop density and service quality for speci c areas of London. e results show that mode choice combinations such as "walk to shop and take the bus back" are not as frequent as sometimes thought, and that bus stop density is more important to older people than attributes describing the quality of bus service, such as service frequency.