It's very accurate for some people, it overestimates for others and underestimates for a few. It overestimates for me, but not hugely.

The thing to do is to trust it - for the most part - for at least a month. During that time, be extra scrupulous about logging your food. Log every bite, weigh every bite. Be obsessive about it. A month is a long enough time to even out the day to day weight variations that can make it hard to figure out calories in vs. calories out. What I mean by trusting your Fitbit "for the most part" is that I think it is reasonable to leave some of your adjustment calories uneaten - as long as you're not aiming for an unrealistic weight loss rate to start with (no more than .5 pound a week for every 25 pounds you have to lose).

After a month, you can go to your Fitbit profile page and it will show you a graph of calories in vs. calories out for the past 30 days and also your average values. Using those average values, you can compute your "theoretical" weight loss for that period. Then compare that to your actual weight loss and, assuming you were really good about logging your food, that will tell you how accurate your Fitbit is for you. Because mine overestimates my calorie burn a little (and because being obsessive about accurate food logging isn't good for my mental health), I have a standard "Quick Add Calories" amount that I include as part of my breakfast every day. I could just try to leave that many calories uneaten at the end of every day, but I find that for me it works better this way. The calories in vs. calories out math still works the way I do it.

Not really. I cant get my head around earning 1000 extra calories for simple walking... I shaved off an inch from my height and stride length hoping it wouldn't give me so many calories, but I still get a boat load every day.

My weight hasn't budged much, either my food logging is woefully inaccurate which is unlikely or my fitbit is over estimating.

1000 extra calories does sound like a lot. Maybe you should bump up your MFP activity level? I had mine set too low and it made every really strange.

Now I only get adjustments if I seriously try and at the most about 400 calories after 5 miles.

Since I just switched from a Flex to a ChargeHR, I don't trust it yet. After 2 days, I am finding my earned cals to be less than before for the same exercises and walking the same number of steps. I think the HRM must be the reason for that. The flex was impressed with all my steps, the ChargeHR not so much. I will say that I am not wearing the two at once but comparing old and new data, so I can't make a direct comparison.

This is precisely why i will never get a HR monitor. As the only exercise I do is walking, and not at Olympic athlete speed either

@L_Master can you let us know your results with the your HR switched off? I'd be interested to see if you get a more realistic burn. BTW we are exactly the same weight and height, except I'm nearly 20 years older . You'd have the metabolism of a race horse

heybales wrote: »

Calories based on pace and weight formula are the most accurate - more than HRM even.
http://community.myfitnesspal.com/en/discussion/774337/how-to-test-hrm-for-how-accurate-calorie-burn-is/p1

If there is good reason NOT to trust them - like you see where there are tons of extra steps, or distance is way off, ect.

This part I'm pretty confident with. Distance seems good, steps are damn near perfect outside running or walking and indoors seem well within acceptable margins of error.

heybales wrote: »

Also - the HRM based formula isn't used for under-exercise level activity. Step-based is.
The kicker until the Fitbit knows you after a couple weeks is - where is that level initially to switch modes?

Now THAT is super interesting. If that's how fitbit works, that's really cool. If only they made stuff like this clear right from the front, but they don't mention that on their website or FAQ anywhere clear.

The other possibility that could be playing in is that the HR doesn't seem to get measured quite right when I'm on the move and being a bit fidgity. Checking my pulse it's usually 60s standing and 70s light walking indoors. Fit often will read around there but with a decent number of spikes up to 90s/100s. Same thing if I press the HR button I'll often seen 80s/90s standing.

heybales wrote: »

I'd suggest if you got 334 kcal standing, where Fitbit would normally give BMR level calorie burn for no moving - then you may have some stats that Fitbit is using wrong.

I'll go back in and check this, pretty sure everything is in there correct but I may have made a mistake somewhere.

heybales wrote: »

There is a bug that still crops up - despite the fact of seeing lbs on Fitbit's site, it's using that figure as KG and the energy burn that goes along with that mistake is higher by 2.2 than it normally should be.
It's easy to confirm by looking at the lowest low in daily 5 min graph at sleeping/non-moving time.
That figure / 5 x 1440 should be close to Mifflin BMR you can look up on MFP under Apps - BMR calc.
Sadly that figure is rounded for display on Fitbit - but it still should point out that bug if it's being used.

Doesn't seem to be the case for me. My lowest 15 min period's in the night are 17 kcal. So 5.66666 for 5 min block. Divided by 5 is 1.133333 which x 1440 = 1632. Pretty close to the 1652 given to my by MFP.

I am nervous about the same thing so I changed my mfp to sedentary instead of lightly active since Fitbit accounts for my entire day. Mfp dropped my calories down by 200 so I will see if this makes a difference.

Bbeliever215 wrote: »

I am nervous about the same thing so I changed my mfp to sedentary instead of lightly active since Fitbit accounts for my entire day. Mfp dropped my calories down by 200 so I will see if this makes a difference.

All this means is that your Fitbit daily adjustment will be higher by about 200 calories - it doesn't change anything at the end of the day except giving you a smaller number to plan on and the be surprised with a bigger adjustment.

But it does improve end of day final adjustments. Read second half of FAQ for what that means.

I am having a hard time with this too - just got a Charger HR, but had been using MFP and Pacer app for about a year before that. With Pacer I would end up getting around 600 extra calories a day and I am pretty active - I work out and walk on average 25 000 steps a day. With Fitbit I earn over 1000 extra calories; yesterday it seemed to do something really strange and said I earned almost 3000 calories from a 25-minute workout and walking 35 000 steps (28km). At 5'6" and 120 lbs, there is no way that I burned that many calories. One particular walk was just over an hour long and apparently burned almost 600 calories. If that was true, and Pacer had been off all that time then I would have had a way harder time maintaining my weight, which I have done for a while now. So I am pretty disappointed because I thought Fitbit would be more accurate.
Anyone have any ideas for settings I could change? The lightly active/etc. settings on MFP don't help enough. Even if I set MFP to say I want to lose a lb a week (I am just wanting to maintain my weight), it doesn't seem to lower the calories enough. The problem is that FItbit seems to vastly overestimate how many calories I burn.

Any tracker you wear is only ever an algorithm estimating, and can't directly measure calorie burn. So it's gonna be a guess.

I had a Flex for a while. Now I have a Charge HR. The calories I get from the Charge HR are about 100/day lower than what I got from the Flex for the same activity at the same time. (I even wore both, on different accounts, to compare for a while). I find the more conservative estimate from the Charge HR is closer to accurate for me, though there are certain types of non-step based activities that no Fitbit can properly measure.

The only way to know how accurate it is for you is to try it for a while and see.