FRET experiments enable studies of the chemical and physical properties of individual molecules, which has long been a dream of chemists. However, these modem experimental techniques are still limited by the lack of information about the dynamic behavior of the fluorescent labels as well as by the use of dipole-dipole approximation even at short donor-to-acceptor distances. Our results help to suggest that these assumptions need to be carefully considered when designing experiments. We show that at short donor-acceptor separation, dipole-dipole approximation breaks down and Forster theory fails and cannot be used to obtain correct distances. We also explicitly demonstrate that dyes' linkers allow for a lot of flexibility in the fluorescent label orientation and position resulting in distances much shorter than assumed earlier.