Amended panarthropod phylogeny. Strict consensus of 57 (left) and 147 (right) most parsimonious trees under implied weights (k = 4), + denotes extant onychophoran taxa. Arthropod and tardigrade total-groups collapsed for clarity, as they are unchanged from the results presented by Yang et al. . Left-hand topology derived from modifying the character coding for Helenodora inopinata (=Ilyodes) of Yang et al.  to reflect the new observations herein. Right-hand topology derived in the same way but with characters whose absence is inferred from taphonomic analysis coded as ambiguous. For each dataset, the consensus topology for total-group Onychophora was stable with all but the most stringent concavity values (equal weights and implied weights with concavity constants in the range of k = 4 to ∞). Under stricter homoplasy penalization (k < 4) the position of Aysheaia and Onychodictyon ferox are unstable, reducing the resolution, but the position of H. inopinata relative to crown-Onychophora is unchanged. Numbers on nodes indicate node-support based on symmetric resampling, under equal (above) and implied weights (below, k = 4). The results shown are difference in frequencies, given as ‘GC’ values (for Group present/Contradicted ). For clades recovered in both analyses, their support is comparable. When taphonomic evidence is used to inform character coding H. inopinata is resolved as a stem-onychophoran, i.e. clearly outside of crown-onychophora, contra Yang et al. . If taphonomic evidence is not used the position of H. inopinata is ambiguous and much of the resolution within total-group Onychophora is lost, demonstrating the utility of experimental and comparative taphonomy in phylogenetic reconstruction.