The risks posed by a range of acoustic scientific instruments were assessed by the construction of matrices of scale and likelihood. We recognized six levels of impact ranging from none or short term, minimal behavioural response (Level 1) to multiple injuries and fatalities and/or compromised populations (Level 6) and six levels of likelihood ranging from “Expected in almost all instances” (Level 1) to “cannot see how it could happen” (Level 6). Typical scientific instruments ranging from acoustic releases to large air gun arrays were assessed. To provide a perspective for the risks of scientific operations, other activities were also ranked. These included large chemical explosions, submarine detection sonars implicated in some mass strandings of cetaceans and normal Antarctic shipping activities. The conclusion reached was that most scientific instruments pose a similar or lower risk than normal shipping operations. High source-level equipment poses some risk to individual animals' hearing and so should be mitigated. Likewise, survey planning should be designed to avoid trapping animals in narrow, constricted sea ways. Long term, cumulative impacts are still difficult to detect in areas with greater anthropogenic noise than the Antarctic but we concluded that any possible long term impacts should be mitigated by maintaining the low levels of activity using high source-level equipment through data sharing and survey planning.