It could well be the lab, but before you blame them totally, you also need to factor in the human visual system, which is rather lax at recording colour.
We see things as the colour we "know" they are, factoring out lighting. This is usefull as it means we still know that the large orange and black thing, which looks red and back in dusk lighting is still a tiger. The most comon photographic example of this is tungsten lighting Vs daylight. We don't see light bulbs as orange but the film does. The film is "right", but not what we want, so we correct for it.
Unless you're scanner, and monitor are calibrated, then there's no reason to sugguest that it's any more "right" than the lab - just more as you'd like it to be. If you're printing out from a scanned image, then you're even more likely to be getting random colours.
I'd be cautios of calling either version "right" - it's more likely that the monitor is simply closer to your perfered vision, than the lab print. This re-enforces the do your own darkroom stuff argument - that way you get the version you want. When we print b&w we don't produce one print and call it "right" - we produce a selection, and pick the best, or prefered. Thats a value judgement. Why should colour printing be any different?
Of course it's also possible the lab screwed up...
For some REAL fun re colour vision try reading: http://www.wendycarlos.com/colorvis/color.html
(and yes it IS that Wendy Carlos).