The genetic transformation of plastids of higher plants has developed into a powerful approach for both basic research and biotechnology. Due to the high copy number of the plastid genome per plastid and per cell, repeated cycles of shoot regeneration under conditions selective for the modified plastid chromosome are required to obtain transformants entirely lacking wild-type plastid genomes. The presence of promiscuous plastid DNA in nuclear and/or mitochondrial genomes that generally contaminate even gradient-purified plastid fractions reduces the applicability of the highly sensitive PCR approach to monitor the absence of residual wild-type plastid chromosomes in transformed lines. It is therefore difficult, or even impossible, to assess reliably the hetero- or homoplastomic state of plastid transformants in this manner. By analysing wild-type and transplastomic mutants of tobacco, we demonstrate that separation of plastid chromosomes isolated from gradient-purified plastid fractions by pulsed-field gel electrophoresis can overcome the problem of (co)amplification of interfering promiscuous plastid DNA. PCR analyses with primers specific for plastid, mitochondrial and nuclear genes reveal an impressive purity of such plastid DNA fractions at a detection limit of less than one wild-type plastid chromosome copy per ten transplastomic cells.