Admissibility 

Whether an item of evidence may be put before a court is not answered by a single test. It is answered by working through a sequence of questions, each of which can end the enquiry before the next is reached. Understanding that sequence — and not just the individual rules within it — is what separates a competent evidence answer from a list of disconnected rules.

Step one: relevance

Every admissibility analysis begins with relevance, addressed in full elsewhere in this module. Evidence must be capable of rationally affecting the assessment of the probability of a fact in issue. If it is not relevant, the analysis stops there: no exception, no expertise, and no absence of prejudice can rescue irrelevant evidence — s 56 of the Evidence Act 1995 (Cth) and its uniform counterparts admit relevant evidence except as otherwise provided, and exclude irrelevant evidence outright.² Relevance is necessary but never sufficient, which is why the analysis continues even once it is satisfied.

Step two: the exclusionary rules

Relevant evidence is then tested against a set of rules that exclude categories of evidence regarded as unreliable, unfairly one-sided, or contrary to a protected interest, unless a recognised exception applies. The major exclusionary rules are:

• the hearsay rule — excluding previous representations tendered to prove the truth of what they assert, because the maker cannot be cross-examined;
• the opinion rule — excluding opinion evidence in favour of evidence of fact, subject to the expert and lay opinion exceptions;
• the tendency and coincidence rules — restricting evidence used to show a person's disposition or the improbability of coincidence, because of the risk of prejudicial propensity reasoning; and
• privilege — protecting certain confidential communications and interests from compelled disclosure regardless of relevance.

Each of these operates as a general rule of exclusion with defined exceptions, and each is addressed in its own article in this module. An admissibility analysis has to check the evidence against each rule that is realistically engaged on the facts — a document may raise a hearsay question and a privilege question at the same time.

Step three: discretionary and mandatory exclusion

Evidence that survives relevance and every applicable exclusionary rule is still not automatically in. Courts retain, and in some circumstances are required to exercise, a power to exclude evidence that is technically admissible where its probative value is outweighed by the danger of unfair prejudice to a party, or by the risk that it will mislead, confuse, or consume a disproportionate amount of time. In criminal proceedings this concern is particularly acute where evidence, though logically probative, risks being given more weight than it rationally deserves by a jury — for example, because of its emotional impact rather than its true bearing on the facts in issue. Some jurisdictions frame this as a discretion; others frame aspects of it as a duty to exclude in defined circumstances. Evidence obtained improperly or illegally can also attract exclusion on public policy grounds, per Bunning v Cross (1978) 141 CLR 54,¹ balancing the desirability of admitting the evidence against the undesirability of admitting evidence obtained in that way.

In a uniform Evidence Act jurisdiction that discretion is no longer the operative rule. Kadir v The Queen (2020) 267 CLR 109³ traces s 138 to the ALRC's proposal for "a modified form of the common law exclusionary public policy discretion articulated in Bunning v Cross" ([11], Kiefel CJ, Bell, Keane, Nettle and Edelman JJ). The modification matters most in where the onus sits: under s 138(1) the evidence "is not to be admitted unless the desirability of admitting" it outweighs the undesirability, so exclusion is the default and the tendering party must justify admission — the reverse of the common law position.

Where a statute itself prohibits use of an unlawfully made recording, the question may be one of statutory construction rather than discretion. Potter (A Pseudonym) v The King [2026] HCA 25 (5 August 2026)⁴ concerned a wife's covert recording of her husband admitting earlier sexual assaults, tendered at his trial in South Australia, which is not a uniform Evidence Act jurisdiction. Under the Surveillance Devices Act 2016 (SA), a party to a private conversation may not record it with a listening device (s 4(1)(b)) unless an exception applies. The exceptions available to a party who records the conversation include that all principal parties consent (s 4(2)(a)(i)) or that the use of the device "is reasonably necessary for the protection of the lawful interests of that person", meaning the recording party (s 4(2)(a)(ii)). Section 12(1) separately prohibits knowingly using, communicating or publishing material derived from a use in contravention of Part 2, subject to the exceptions in s 12(2).⁵ The joint reasons (Gageler CJ, Gordon, Steward, Jagot and Beech-Jones JJ) said that if the recording contravened s 4(1), its use at trial was prohibited by s 12(1) and that, "reflecting the parties' common position in this Court, no question of any discretionary exclusion (or inclusion) of the evidence of the covert recording should have arisen" ([14]). The respondent had conceded that the Court of Appeal erred in treating admissibility as governed by the public policy discretion in Bunning v Cross, and the joint reasons said that concession was correctly made. The only issue on the appeal, however, was whether s 4(2)(a)(ii) was satisfied. The Court held that it was, so the recording was not prohibited and was properly admitted; the effect of a statutory prohibition on the discretion was common ground, not the question the appeal decided.

Why the sequence matters

Treating admissibility as a single test invites two common errors: concluding evidence is admissible because it is obviously relevant, without checking the exclusionary rules; and conversely, spending an entire answer on a single exclusionary rule while forgetting that discretionary exclusion remains live even if that rule is satisfied. A disciplined analysis walks through relevance, then the exclusionary rules genuinely raised by the facts, then discretionary and mandatory exclusion, in that order, for each item of evidence in turn.

The jurisdictional dimension

The content and even the existence of some of these steps varies with jurisdiction. The uniform evidence law, based on the Evidence Act 1995 (Cth) and Evidence Act 1995 (NSW), applies with local variations in the Commonwealth, New South Wales, Victoria, Tasmania, the Australian Capital Territory and the Northern Territory. Queensland, South Australia and Western Australia have not adopted it, and instead apply their own evidence law, closer to the pre-existing common law and their own separate Evidence Acts. The exclusionary rules described above exist in some form across Australian evidence law, but their precise formulation, the shape of their exceptions, and the balancing language used for discretionary exclusion differ between the uniform scheme and the three non-uniform states. A well-founded admissibility answer names the governing jurisdiction before applying rule detail.

Applying this in a problem question

1. Work through each item of evidence separately rather than the case as a whole.
2. Apply relevance first, and stop there if it is not satisfied.
3. Identify which exclusionary rules are genuinely raised on the facts, and apply each in turn, checking for a recognised exception.
4. Even where the evidence clears every exclusionary rule, consider discretionary or mandatory exclusion for unfair prejudice, confusion, or impropriety in how it was obtained.
5. State the jurisdiction and whether the uniform evidence law or a non-uniform state's own scheme governs, since this affects the detail of steps two and three.