Empirical data on the transfer and persistence of trace DNA are crucial to the evaluation of forensic DNA evidence. This evaluation can be complicated by the occurrence of indirect DNA transfer; the possibility of which is well established, but research into such transfer is often focussed on unrealistic situations, e.g. handling of DNA-free items after participants have shaken hands for 1-2min. To simulate more realistic scenarios, this study investigated the deposition and persistence of both directly- and indirectly-transferred DNA on knives that had been artificially set up as 'regularly-used'. Each knife was handled in a prescribed manner by a specific participant over two consecutive days to simulate regular use. Each participant then shook hands for 10s with a fellow volunteer and immediately stabbed one of their knives into a foam block repeatedly for 60s. DNA was recovered by mini-taping from triplicate sets of knife handles from four pairings of volunteers after regular use, and at one hour, one day and one week after the handshaking and stabbing events. Total amounts of DNA recovered from the knives, regularly used by a single person, varied among individuals; one volunteer consistently deposited significantly greater amounts than the others, whilst another volunteer did not always leave complete profiles. DNA attributed to the regular user persisted for at least a week, declining with increasing time between DNA deposition and recovery. Non-donor DNA was co-deposited at <5% of the profiles recovered, except for one volunteer, who consistently left DNA from their romantic partner on their knives at ∼25% and ∼11% of the profiles before and after the handshaking and stabbing events, respectively. In three pairings of volunteers, after the handshaking and stabbing events, alleles that could be attributed to the respective handshakers' profiles were detected as partial minor profiles, equating to ∼10% of the profiles recovered. For the fourth pairing of volunteers, only complete single-source DNA profiles matching the regular user's profile were recovered. However, it is important to note that, when indirectly-transferred handshaker DNA was detected, it declined with increasing time between DNA deposition and recovery. These data provide an initial insight into the detection and persistence of directly- and indirectly-transferred DNA that extend the data already available on forensic DNA transfer. The results herein suggest that the sooner an item is sampled after an offence has occurred, the greater the chance of recovering indirectly-transferred DNA, which has implications for forensic reconstructions.