Not as many takers on this thread as I thought there might be. :o
FWIW:
Wearing an HRM to bed is not the best method. For a start they can be uncomfortable!
More importantly your resting HR is not a single reading (ie. the single lowest HR recorded over a night's sleep). Such a reading is more likely to well outside the mean. Palewin's post mentioned that using a whole minute's worth of readings might not be sufficient as the results still might be skewed too far from the true mean.
Further, your heart rate can drop to very low levels during deep sleep. In fact, there is a point in the wee hours of the morning where most people sleep so deeply that their heartrate is below their true resting heartrate.
As for taking the reading before going to bed, Aussie531 has already highlighted a few problems (eg. fatigue, dehydration, influences such as alcohol, stress, etc) with that method. But the main problem with it is that your body hasn't even started recovering! If you trace your heartrate during overnight recovery you see nice little saw-tooth ramps as your body mends itself and the heart rate settles down. The difference in HR between going to bed and waking can be significant. Especially if you're watching your wife doze off to sleep and you mind (or hands) begin to wander. 😉
Aussie531 also pointed out a common waking phenomenon: high HR caused by a fight or flight response to waking. My HR always stays high for up to 5 mins after getting up and turning off the alarm clock. That's why I rest for a few minutes before starting the reading.
And finally, Palewin mentioned that I should average the whole five minutes. He is absolutely on the money when he says that it shouldn't be a game of finding the absolute minimum. But by the same token you want a true indication, so you need to eliminate error where possible. Averaging does this quite well, but all too often I see steep ramps at the start and end of a 5 minute reading that are obviously abberrant data.
So I'm gunning for a 3/5 method: Rest five minutes after waking (without falling back to sleep!), read for five minutes then eliminate the first and last minutes so that any fidgetting and fussing doesn't influence the reading.
Any other thoughts?