Using a gamma-ray detector and a television camera system for synchrotron light, high-energy bremsstrahlung and horizontal growth of the synchrotron light source were observed when sudden decrease in the electron-beam lifetime occurred due to dust trapping in the electron beam. Two types of beam current losses were found; one was a continuous beam current loss, and the other was a short-term beam current loss. High-energy bremsstrahlung at a location was observed in a short time and after that, the bremsstrahlung was not detected in spite of the occurrence of dust trapping phenomena. The fact suggests motions of the trapped dust particles in the longitudinal directions. Materials collected in the beam chamber are dust particles from ion pumps and dust particles made during the beam chamber processing for welding. Most of the collected dust particles were less than 2 mm in size and surfaces of some dust particles were melted with the electron beam. Simple analysis was carried out for the conditions necessary for a dust particle to be trapped, for motions of the trapped dust particle, and for interactions between the trapped dust particle and the electron beam. The analysis showed that a dust particle less than 3 mm in size, made of Al, can be trapped and that the trapped dust particle can move in the vertical and longitudinal directions. The analysis also suggested that a dust particle in size of about 2 mm can be continuously trapped around the electron beam without being destroyed by the electron beam. Furthermore, the analysis explained the difference between the two types of beam current losses observed in the ring. Experiments which simulate the electron beam using a Cu wire in an evacuated beam chamber show that a dust particle (less than 70 μm) is trapped sufficiently. The experiments also coincide with theory for an attractive force acting to a conducting small particle. The calculated electric field of the electron beam and the calculated electric charge of dust particles given through the photoelectric effect in the TRISTAN accumulation ring are 100 times and 104–106 times higher than those of the simulated experiments, respectively. In the ring, the attractive force caused with the average electric field and with the expected charge is 10–103 times larger than that of the simulated experiments. Therefore, a dust particle (less than 2 mm) can be trapped sufficiently. An electrostatic dust collector using an electron beam and an electrostatic force are effective in removing all of the sample dust particles in the test chamber for the simulated experiments. A method to remove trapped dust particles using electrostatic electrodes is also discussed. It is expected that such electrodes can be useful for trapped dust particles moving in a longitudinal direction.