Short answer, don't do it*.

Longer answer, yes, I've tried this, and no, it doesn't work well.

260/230 is primarily a measure of guanidium salt contamination (phenol absorbs mostly at 270), and guanidium salts, being chaotropic agents, are a bad thing for downstream reactions like cDNA synthesis.

In my experience, anything with a 260/230 ratio below about 1.7 gives increasingly poor data, and it drops off pretty fast, too. 0.3 is literally unusable.

*However, the important thing to remember here is that 260/230 is a ratio, not an absolute. We use ratios because this approach factors in the RNA concentration, and when it comes to cDNA synthesis, we'll usually be trying to use a similar amount of RNA (total), not a similar volume. If you have LOADS of RNA, a poor 260/230 means you also have LOADS of salt. If you have very little RNA, a poor 260/230 means you also have...a small amount of salt. In either case, if you're trying to set up a reaction with "X micrograms of RNA", you will be adding a similar amount of salt.

In your case, this may not be a problem: with very low RNA yields, the 260/230 ratio becomes counterproductive, because even small amounts of salt contamination will produce a very poor ratio. Not because the guanidium salt level is high, but because the RNA level is low.

If you were to (somehow) use 1-2ug of RNA in a reaction, then this might be problematic, but...you probably can't, because you don't have enough RNA in the first place. Your reaction conditions are limited by total volume, and per unit volume, you have less RNA.

If you have less RNA, you also have less salt.

A cDNA synthesis reaction that absolutely will not work with 1ug of RNA with a 260/230 ratio of 0.7...will probably work absolutely fine if you only use 100ng of RNA (because you only have 100ng of RNA).

So...in this instance, what I would do is to sort of...reverse engineer the ratio: take all your samples and divide the 260/230 ratio by the actual [RNA]. If you get mostly the same number, that implies your level of salt contamination is a feature rather than a bug, and that those samples with really low ratios are also those samples with really low [RNA] (i.e. your salt contam level is consistent, it's your [RNA] that varies). After all, you will always have _some_ salt contamination, and this simply becomes more obvious as your RNA yields get lower.

And then, really: I'd just pick a sample and have a go. By the sound of it, you don't have much RNA, so you may end up using only 100-200ng of RNA per reaction. If this is the case, then this might be fine, because your guanidium salt level will be commensurately lower.

If that one sample gives you data you trust, then do the same for the rest.

I have had pretty reliable success in situations where I used what I had, even when the ratio was poor, simple because what I had was "not very much".

John Hildyard Thank you so much for the detailed explanation, it makes so much sense! Yes I was planing on using between 100-200ng of RNA, since my yield was low and as you mentioned im limited by volume. I was planing on doing only a 1/10 cDNA dilution after RT-PCR since i used less RNA, would that be ok?

1/10 should be fine.

You can always test with a dilution series: use 1ul of neat cDNA, 1ul of 1:2 diluted, 1ul of 1:4 diluted etc, down to maybe 1:64 or 1:128 diluted.

You should see a nice linear relationship between log(dilution) and Ct somewhere in there: it might be non-linear at high concentrations and low concentrations (due to PCR inhibition and stochastic template numbers, respectively), but as long as your data falls on a linear scale over the ranges you're interested in, you're good to go.