Depot injectable testosterone undecanoate (TU) was introduced in the last decade but its safety experience in clinical practice is not well defined. We surveyed prospectively TU injections given in the Andrology Department, Concord Hospital over 3.5 years (3,022 injections given to 345 patients) to estimate the incidence of (a) immediate cough/syncope related to oil vehicle microembolisation, (b) self-reported post-injection hematoma in patients taking antiplatelet &/or anticoagulant drugs and (c) the prevalence of secondary polycythaemia during treatment. Immediate cough/syncope was observed after 56 injections (66% mild, 16% severe; 40% before injection completed). The incidence was 19 (95 % CI 14-24) per 1000 injections with patients displaying an excess of two or more episodes per patient (vs one or none, p<0.001 Poisson distribution) but the incidence rate was no different between 3 experienced nurse injectors. No post-injection hematoma was reported (upper 95% confidence limit <0.1%) including after 269 injections to patients taking antiplatelet, anticoagulant or both drugs, of whom 56 did not withhold drugs prior to TU administration (upper 95% CL 5.4%). Mean PCV during treatment (n=345) was 0.47 ± 0.05 (SD) (range and quartiles 0.31, 0.44, 0.47, 0.50, 0.59) with 77 (22%) >0.50, 39 (11%) >0.52 and 13 (4%) >0.54. We conclude that while TU injection is generally safe and well tolerated, the incidence of immediate cough/syncope is 1.9% (1.4-2.4%) following injections by experienced nurses with recurrence among patients experiencing an episode more likely than by chance. Most cough/syncope was mild with nearly half occurring during slow deep im injection. Post-injection hematoma was not observed, even among men continuing to take anticoagulant &/or antiplatelet drugs. Mild polycythemia is relatively common but rarely required specific treatment other than optimising inter-injection interval.