Forty seven potato virus A (PVA) isolates from Europe, Australia and South America’s Andean region were subjected to High Throughput Sequencing, and 46 complete genomes from Europe (9), Australia (2) and the Andes (35) obtained. These and 17 other genomes gave alignments of 63 open reading frames 9180 nucleotides long; nine were recombinants. The non-recombinants formed three tightly clustered, almost equidistant phylogroups; A comprised 14 Peruvian isolates, W comprised 37 from Peru, Argentina and elsewhere, and T contained three from New Zealand tamarillo. When five isolates were inoculated to a potato cultivar differential, three strain groups (= pathotypes) unrelated to phylogenetic groupings were recognized. No temporal signal was detected amongst the dated non-recombinant sequences, but PVA and potato virus Y (PVY) are from related lineages and ecologically similar, therefore ‘relative dating’ was obtained using a single maximum likelihood phylogeny of PVA and PVY sequences and PVY’s well-supported 157 CE ‘time to most common recent ancestor’. The PVA datings obtained were supported by several independent historical co-incidences. The PVA and PVY populations apparently arose in the Andes about 18 centuries ago, and were taken to Europe during the Columbian exchange, radiating there after the mid-19th century potato late blight pandemic. PVA’s phylogroup A population diverged more recently in the Andean region, probably after new cultivars were bred locally using Solanum tuberosum spp. tuberosum as a parent. Such cultivars became widely grown, and apparently generated the A x W phylogroup recombinants. Phylogroup A, and its interphylogroup recombinants, might pose a biosecurity risk.