The present research sought to assemble the plastomes of 42 species in the Apocynaceae, tribe Marsdenieae. Fast-Plast, a chloroplast assembly pipeline, was used to assemble the plastomes, with alignment to reference plastomes (Asclepias syriaca and Nerium oleander) and manual editing of the alignments carried out in Geneious. Marsdenieae plastomes proved difficult to assemble, due to presequencing, sequencing, or assembly errors or biological realities. Only 35.7% of the species of Marsdenieae included assembled at all, and no finalized contigs were produced for any species. In most taxa, not all of the contigs aligned with either reference, and nearly all alignments had gaps between contigs. Fast-Plast-assembled contigs and raw reads exhibited many mismatches and gaps when alignments were visually-inspected, the result of errors or biological realities. However, when four metrics—number of contigs assembled, percent of assembled contigs that aligned to each reference plastome, percent coverage of contigs aligned to each reference sequence, and NG50—were considered, most species were deemed to have assembled well, suggesting that these metrics are insufficient to assess assembly quality. Although alignments to the N. oleander plastome were slightly better, there was no significant difference between these alignments and those to the Asclepias syriaca plastome. When raw reads were aligned to reference plastomes, a pattern of high coverage, accompanied by many gaps and mismatches, suggested that greater genetic distance from a reference sequence may be responsible for at least some of the poor alignment outcomes. Suggestions for improving the alignments and assemblies are given, with an emphasis on possible reasons for the high numbers of variable regions, including assembly or alignment errors, sequencing problems, and taxon-specific chloroplast structure.