SEE THE UPDATE at the end of the review.I love automatics and this is my first Seiko auto (the others have all been Swiss in origin and mid-level, but not "luxury", and all with Swiss ETA 2824-2 cal either stock or modified). I was looking for a "beater" everyday watch that wasn't flashy in anyway, versatile, inexpensive, from a reputable name in horology, had a basic military style, was at least 100m water resistant, preferably with a sapphire crystal, display back would be nice and one that would go well on NATO straps. This Seiko hit ALL of the points EXCEPT the crystal, which is Seiko's proprietary Hardlex. From what I've read, it's a step above the standard mineral crystal and yet still below sapphire. Fine, it's not a luxury watch and for $109, I can live with Hardlex.I'd like to address points I've seen in many reviews here on Amazon, along in this review to help others out. A lot, by no fault of the reviewer, are really based out of not understanding mechanical/automatic watches in general.DESIGN: Basic, no nonsense, military. Will not go out of style, this has been around since the 1930's.CASE: Bead-blasted stainless. A nice matte finish that doesn't give off a lot of reflection and is consistent with military style. The back is polished, so is the area between the lugs. A solid case. The display back is a nice touch. 42mm is a bit big for this watch, I think. I'd rather see it at 40mm and it would still be a nice step up from the tiny little 809 at 36mm.FACE: A bit busy for my taste, but I can live with it. The info that is needed is easy to read and the luminosity ("Lume") is VERY good. What I'd change is to remove the "Sports" designation from under the Seiko 5 logo, it's just dumb and not needed. I'd also like to see the minute and second hands be a bit longer and reach onto the chapter ring. Small things, not deal breakers. The LUME, wow!MOVEMENT: Here's where people get confused and frustrated and complain.#1 You have to remember that this is an inexpensive watch. You cannot get around that fact. In an autowinder, you are not getting Swiss chronometer certification here (LOL). Additionally, quartz movements are almost always more accurate than mechanical. That's not the point (BTW, I dislike quartz personally).#2 This is powered, like all autowinders, by a mainspring that must be wound. When it's FULLY wound (or close) the stored energy is what runs the watch. The movement of your arm, or shaking it gently in a looping motion, will wind it. Give it a good 100 to start with when you first get it, then set it and put it on. See #3.#3 This watch does not offer a manual winder via the crown like most autowinders do. For the price of this watch, I'm fine with it. Just know it going in and then don't complain about it when it acts as designed.#4 This watch does not "hack". That means it does not stop running when you pull the crown out to set it. Again, a bit unusual, but for the price it's fine and it keeps the cost down. This really isn't a big deal. If you need 100% accuracy and to hack the time, a $100 auto isn't what you need anyway. For what it is, this is fine.#5 ALL autowinder watches have a power reserve, that's the stored power in the main spring. This one should be about 35-40 hours if FULLY wound. If not it will be less. Mine has been fine overnight, never an issue. My ETA 2824-2's are in the 40-50 hour range, so this isn't too bad. You have to really get up in price for the reserve to be measured in many days or weeks (Think >$10k watches). If you don't want to deal with keeping it moving / winding, mechanicals are NOT for you. Go get a quartz with a batter, solar or kinetic and be happy.ACCURACY: While part of the movement, I want this to stand on it's own. Per Seiko, the accuracy is considered within operational limits at -20 seconds or + 45 seconds per day. It can be regulated to be more accurate by a watchmaker OR someone who knows what they are doing at home. Most users find the 5 series to run about +/- 10 seconds or so. My Swiss ETA 2824-2's run an average of about +/- 1-3 seconds per day, but they cost 5-10x more. For the money, I can live with the 5's accuracy. For more info, please go to Seiko's website and read their FAQ on mechanical watches. I suspect we'll see about a 3-5 second per day of accuracy, maybe, at the worst 5-10 seconds.BOTTOMLINE: This is a fantastic value for an autowinder mechanical watch. I don't believe there is any better value out there than the 5 series if you are looking under $250. Even with the limitations I mention, I still give it a 4 star rating. I'm not deducting stars based on unrealistic expectations of a $109 purchase. If this were $500+ watch, it would be 2 maybe 3 stars at most....maybe. But in it's intended market, this is a stellar value.UPDATE: I set my most accurate ETA 2824-2 movement in a Swiss watch vs. the Seiko to see how it holds up. It has been 48 hours since I started this. The ETA movement runs about 2 seconds slow per day, it was just cleaned and lubricated and still has the same accuracy as before. Thankfully. So how has the Seiko held up? In 48 hours, the ETA movement was set as the "standard" to compare it against the Seiko. I stopped it running (hack) and started it up again when the Seiko was at the top of hour down to the second. They were both fully wound. The Seiko by wearing it all day on a NATO strap, plus some "hand winding" before by moving it in large circular motions about 100 times to get it to at least 75-80%. The ETA movement was hand wound to about the same. The ETA sat on the desk all day and the Seiko was worn. The SEIKO has lost 44 seconds in two days compared to the ETA. The ETA lost 3 seconds compared to a timer on my iPhone.That makes the total time for SEIKO lost at 47 seconds in 48 hours (23.5 seconds per day / ~1 second per hour). The ETA lost ~0.125 seconds per hour. Over a week, if this rate continues, the Seiko will have lost 2.74 minutes and the ETA movement 6 seconds. That may not seem huge, but over a 30 day period, the ETA will be ~1.5 minutes lost in the month. The Seiko? At a rate of 23.5 sec/day that would be a loss of ~11.75 minutes.The Seiko loss is > than the specified range of -20 --> +45/day. A watch that is off by this much is now making me question it. I'll run the test again. If it's better, I'll post it here. If not, I won't and I'll let this new star rating stand. I have dropped it from 4 stars to 3 stars. If the new test is also bad, I'm not taking it to the watchmaker to have the timing adjusted and spending probably around $75 to have it done on a $109 watch. This will be returned quickly if the test fails again. I was not expecting this at all.
The reviews of the 7S36 Seiko movement in this watch fall into three main categories:1. "I don't care how well it keeps the time"2. "Mine only loses a few seconds a day"3. "The time is all over the place, it loses a minute every day or two"The first watch I was sent fell into category 3. The second watch was near perfect - category 2! Here's what I noticed was different, something any user can look for when they are evaluating the performance of the 7S36 movement in this watch. There is a regulator in the movement, although sometimes you have to tip the watch so that the winding weight stops you seeing the regulator! This photograph was taken through the plastic back of my good watch, looking at the regulator with a microscope. The regulator is actually small, but you can see it clearly with a x10 magnifying glass. There is no need to remove the back or do anything to the mechanism!You can see that the arrow near the + and - signs is approximately at the middle (0) position. This is where it should be. It means that when the watch was adjusted in the factory every part of the mechanism was within tolerance, the regulator did not have to be adjusted very far from its nominal position to make the timekeeping accurate. The watch I received which was keeping bad time had its arrow point right at the maximum negative mark, meaning that a lot of compensation had to be applied to get the watch keeping roughly correct time. It was running about 25 seconds slow a day, but would surge sometimes to half a minute fast. Those positive surges are perhaps why the regulator was on hard negative, even though the watch was running very slow? I don't know that answer, I am afraid. Perhaps the time-regulating robot in the factory just had a bad day. Who knows?Anyway, it seems that by taking a quick look through the transparent back of the watch you can get a good guide as to the watch's factory timekeeping accuracy.