A total of 557 newborn piglets were used to compare eight identification devices, including one plastic ear tag as a control (C, n = 348) and two types of electronic ear tags (E1, n = 106; and E2, n = 103), and five types of injectable transponders (n = 557); small 12-mm (D12, n = 116; and S12, n = 110), medium 23-mm (T23, n = 108), and large (32-mm, T32, n = 115; and 34-mm, S34, n = 108). Injections were made s.c. in the auricle base (n = 248) and intraperitoneally (n = 309) using a new technique. All piglets were identified with two devices, but using electronic ear tags in conjunction with injection in the auricle was avoided on the same pig. Readability of devices was checked during fattening (until 110 kg BW) and slaughtering. On-farm losses were lower for control than for electronic ear tags (C = 1.1%; E1 = 8.8%; and E2 = 44.9%; P < 0.01); the latter also suffered electronic failures (E1 = 5.5%; and E2 = 55.1%; P < 0.001). On-farm losses of transponders injected in the auricle base were greater in large (S34 = 72.5%; and T32 = 46.3%; P < 0.05) than in small transponders (S12 = 19.4%; and D12 = 17.1%), but T23 (29.8%) only differed from S34. Transponder size did not affect on-farm losses for intraperitoneal injections in which only one loss was recorded (0.4%). All ear tags had similar losses during transportation to the slaughterhouse (1.2%), but no losses were observed in injectables. Slaughtering losses did not differ between ear tags (C = 11.2%; and E1 = 6.4%), but apart from losses, 12.8% of E1 failed electronically. Injection site affected losses and breakages during slaughtering (auricle base = 6.4%; and intraperitoneal = 0%), but recovery time did not significantly differ (auricle base = 28.6 s; and intraperitoneal = 18.9 s). Transponders in the auricle base were recovered by sight (30.2%), palpation (27.4%), or by cutting (42.5%). Intraperitoneal transponders were mainly recovered loose in the abdominal cavity (81.4%), whereas 18.6% fell on the floor. As a result, traceability varied significantly (P < 0.05) between control (86.7%) and electronic ear tags (0 to 68.1%) and injectable transponders, with the auricle base (17.8 to 75.0%) having lower values than intraperitoneal (98 to 100%). Intraperitoneal injection was a very effective tool for piglet identification and traceability, ensuring the transfer of information from farm to slaughterhouse. To warrant the use of this technique in practice, transponder recovery requires further investigation.