We may never know how life came to pass (from a scientific basis). However, the fossil record is very clear of a progression from simple life billions of years ago to progressively more complex life, so evolution has occurred. One can fairly debate the mechanism for this happening. But one can state that that progression exists with quite a bit of certainty.
Well, it's not that clearcut. "The fossil record" is the object of much veneration, but really, it's somewhat circular in its reasoning. Fossils are dated based upon which rock strata they appear in. Rock strata, in turn, are often dated based upon which fossils appear in them, never mind that the same type of fossils are often found in wildly different types of strata, etc. Even the attempt at an independently-derived date for the strata is fraught with difficulties. Radiometric dating methods are quite inaccurate - largely because of the fact that they rely upon obviously wrong initial assumptions.
What am I talking about? Let's take the case of using K-Ar dating. K40 decays to Ar40 via beta emission, with a half-life of 1.248 billion years. Sounds good, right? I mean, with that long of a half-life, this method should be perfect for dating rocks - thought to be in many cases billions of years old - which contain potassium (a fairly common element in the earth's crust). Problem is, to date something by measuring the conversion of K to Ar, you have to know how much Ar was there to start with. So, rockdaterologists operate with the assumption that when a rock containing K was formed via volcanic processes, it did so with zero initial Ar trapped in the interstices. From there, you take a sample of rock, measure the amount of Ar trapped in it versus the %K ideally found in the rock's chemical compound, and voila (pardon my French), you have an amount of K that has decayed to Ar, and you can plug it into your half-life equation and find out the age of the rock.
Problem is, the "no-initial-argon" assumption is invalid. And this has been proven experimentally in the lab. Russian scientists experimentally re-created the formation of certain potassium containing rocks in the lab, using the suitable heats and pressures and whatnot, and found that actually, quite a bit of argon can be trapped in the interstices of rocks while they cool - it doesn't necessarily come bubbling out like was assumed. This, then, suggests that "real" rocks had the same phenomenon occurring when they formed - which throws the dates obtained from this method out of whack - and I mean by hundreds of millions, or even billions of years. It's all about starting assumptions - if those are off, then everything else dependent upon them will be too.