BBC home affairs correspondent
It is entirely fitting that the latest report into the way rapes are investigated and prosecuted should be entitled "Without Consent".
Victim and suspect know each other in eight out of ten rape cases
Because if there is one issue above all else that stands in the way of achieving a higher proportion of rape convictions, it is establishing that sex took place "without consent".
The so-called "stranger" rapes - where a woman is attacked in an alleyway, bushes or mini-cab - are far less common, and not so hard to solve.
The suspect can often be identified from semen, blood or saliva at the crime scene; injuries and marks on the victim tend to support the case that force was used, together with evidence from CCTV or eye-witnesses.
Juries have little trouble convicting in these circumstances.
But in eight out of 10 rape cases, victim and suspect are known to each other.
Increasingly - because of advances in DNA - the suspect can't rely on the defence that sex didn't take place, so the tactic usually employed is that sex was consensual.
Prosecutors fail to exploit "bad character" evidence, the report says
In the majority of cases it's an effective tactic, because with little or no independent evidence to corroborate the victim's account, juries, as they're required to do by law, give the benefit of any doubts they have to the defendant.
What the inspection report makes clear, however, is that in these cases police and prosecutors should redouble their efforts to find evidence which supports what the victim says.
Officers should exploit every scientific avenue they can. A tear in a piece of clothing, for example, might help show that the victim struggled and resisted having sex.
The report also says interviews with suspects need to be carefully planned and more thought needs to go into the way videotaped interviews are conducted with victims - because this may be crucial for the jury in assessing the validity of the allegation.
Another area the prosecution have failed to exploit fully is the use of "bad character" evidence.
The law changed recently to make it easier for details of a defendant's previous convictions and offending behaviour to be put before a jury when it's relevant to the case.
But the report found that this isn't happening enough. In one instance, a rape suspect's history of domestic violence wasn't in the prosecution file when it could have strengthened their case.
Undoubtedly, great strides have been made since the last inspection was carried out five years ago, and with teams of specialist police and prosecutors in place the standard of investigations should improve further.
But there is a limit to what the prosecution side can do.
The reluctance of juries to convict in cases where women are drunk and have initiated or encouraged some intimate contact is a question for society, for the public, for us to address.