The P transposable element invaded the Drosophila melanogaster natural populations between 1950 and 1970, following a horizontal transfer, likely from Drosophila willistoni. We analyzed by genomic DNA PCR and sequencing, the P element status in willistoni group species (D. willistoni, D. equinoxialis, D. paulistorum, D. insularis, D. nebulosa, D. tropicalis) in order to investigate what can be equilibrium states of the P family TEs on their long term in species (still active family or not). We found that potentially autonomous P elements appear to be still present in most of the D. willistoni and D. equinoxialis natural populations, whereas a large proportion of D. paulistorum and D. tropicalis collected lines apparently lack complete P copies. In addition, we have found in all except one more distant species (D. nebulosa) the presence of numerous P-MITEs (~200bp) with conserved sequences to be mobilized in trans and which appear to be identical by descent in all willistoni group species tested. They appear therefore to be old genome components present before species radiation. Strikingly, D. insularis appear to have retained only P-MITEs but no long P copies. We have found that most of the willistoni group species carry a common P element variant, named “protocanonical P element” (pcP), which differs from the canonical P element (cP) of D. melanogaster by three diagnostic marks including a SNP variation in the 3' Transposase Binding Site (TBS) which becomes therefore symmetrical with the 5' TBS, a situation distinctive from that of cP elements. Finally, small RNA sequencing in D. willistoni, D. equinoxialis and D. tropicalis ovaries showed the presence of sense and antisense P-homologous piRNAs, indicating a P repression based on a canonical germline piRNA mechanism. The evolutionary perspectives of the P family in the willistoni group species will be discussed.