For what it's worth, I found this interesting post on the Washington Fly Fishing site about this very subject (resident trout vs. steelhead smolt). I'm not trying to stir any trouble or do any fingerpointing here; I'm just enjoying the exchange of ideas and trying to educate myself. Anyway, here's the post:

Well I imagine you're not too confused about the difference between adult rainbows and juvenile steelhead. The question seems to be how to tell the difference between juvenile (read small) residents, and juvenile steelhead. Unfortunately, the short answer is you can't, not in the field anyway, unless the steelhead is actually "smolting." (That is, preparing to make the transition to saltwater; its scales will be fine and silvery, sluffing off easily when handled.)
What most of us refer to as "smolts" are usually steelhead parr. They generally won't be bigger than 6-8 inches, with rare individuals up to about ten inches. Some notable exceptions can occur in the southern range -- CA and far-southern OR -- where some coastal steelhead streams can go subterraneon at their mouths for years at a time. The juvenile steelhead involutarily "residualize" in the somewhat brackish lagoons, reaching lengths commonly up to 16". Meanwhile of course, the stream is gathering no annual runs of anadromous spawners. Fairly consistent fisheries can develop for these lagoon "trout." When conditions finally allow the mouths to break open, the "trout" almost immediately disappear, and the stream starts gathering genetically distinct anadromous steelhead again.

One of the many things those cases illustrate is the amazing life-history flexibility of steelhead/rainbows. Other posters on this thread have noted the tendency of juvenile steelhead to residualise. But it works in reverse as well. "Resident" rainbows can provide a genetic "bank" for sympatric steelhead populations, when resident rainbows and steelhead spawn together (documented in many areas of the range). Also, resident rainbows (even if both parents were residents) can and will spontaneously smolt and become anadromous (lots of cases all over Washington). This can be an important factor for populations that could use any kind of a demographic leg up (don't suppose there are any of those around here, huh?).

So the long and short of it is that any juvenile rainbow under 8" in an anadromous stream-reach is a possible (or potential) steelhead. The chances go up in rivers with a large ratio of steelhead to resident-rainbows (like most westside Washington streams).

I'll don the professor's cap (he should come with a companion emocon that he's putting to sleep), even though I don't know that I've answered your question. I guess I'd have to say that if you're into nothing but sub-8s with adipose fins on the North-Fork Sky, or the Tolt, or someplace similar, and catching juvenile steelhead bothers you, it wouldn't hurt to move on. Now of course we could get into a whole other discussion of when, where, how, why, and if it's OK to C&R juvenile steelhead. Like everything else, the answers certainly wouldn't be cut and dry.