To good to be true?

That’s about the price I was paying in 1997 when I was in college with no insurance.


I paid $24.88 for each of two vials of Novolin at Walmart on Feb 24 this year. IRC I was paying about $15/vial in 1993 when I first came to the US.

Easy, go in, ask for insulin, get insulin, no questions, though maybe the fact that I was also buying $1000 of G6 sensors indicated that I am a diabetic. Neat online interface to my purchases too:


You can get NovoLog for about $30 a vial now too. Not sure why that isn’t being publicized.


Interesting, haven bought without insurance since 97 and that’s approximately what I was paying per bottle then.

1 Like

Yes, here:, but it requires a prescription in Oregon and I don’t currently have an endo (mine departed for the sun about 6 months ago and I’m still waiting for an appointment from his referral).

I actually asked for “Humalin” and they gave me Novolin. It’s not clear to me whether they are the same; they might have both come from the same DNA but then again they might not and no one seems to be about to fess up as to how many different varieties (primary structure) of human insulin there might be.


This is not the same as novolog.


Eh, I certainly didn’t ask for that, and I would have been weirded out if they offered it; I asked for Humulin (sorry, misspelled it the first time), they gave me Novolin. I.e. two possibly different human insulins.


It does seem that Humulin and Novolin are the same, at least according to PubChem; the searches lead to the same page. Curious; I would expect humans to have multiple versions of insulin, but quite a lot of this stuff is just derived from one human being (who didn’t get paid BTW.)

1 Like

This is just 2 different brand names.

They each make Regular, NPH, etc types of insulin, as “generics”.

1 Like

None of this stuff (well, nothing I could find) addresses the basic issue of how much variation there is in human insulin. This is important because I did ok with pig insulin, which is said to have one amino acid difference from human insulin, poorly on bovine (more differences) and have been experiencing problems with novolog (in the fiasp form - novolog with nicotinamide added) which has 1 difference. Humalog, similarly, has one difference.

So obviously it would be kinda interesting if different humans have different insulin; the analogs are just a single amino acid away from some canonical human insulin and this, supposedly, makes a big difference (I should add that, after reverting to Novolin, I’m not seeing any difference!)

Insulin (human and, I suspect, most mammalian) is 21+30 amino acids. It’s actually two separate chemicals but they come from one larger (longer) chemical part of which is split off to generate insulin. 51 amino acids equals 153 bases.

Genetic variation between pairs of human beings is about 1/1000 bases, so any two humans might be expected to have a 15% chance of having different insulin DNA. That doesn’t mean they have different insulin - amino acids are over coded - 61 of the 64 DNA codes correspond to the 20 amino acids used in proteins so a 15% chance of a base difference corresponds to a 5% change of insulin difference.

That would suggest it is pretty damn certain that there are different versions of human insulin. We can’t assume that every variation is deadly (though some, maybe many, are likely to be) because of the fact that we can sort-of handle pig, bovine and analogs - lots of different insulins. It does seem likely that variation will be less than 5% suggests because the DNA in question is rather important; variation that occurs in DNA we don’t use ends up more likely than variation in DNA we do use, for obvious reasons.

Pharmacogenomics examines how DNA variations affect our response to drugs, but this is something of the inverse; maybe rather than going out making DIY changes to get analog insulins it would be more helpful to look at the naturally occurring changes in the human population and how those different insulins behave. Of course the latter would not be patentable (I hope) so commercial companies won’t do it.

Below is a post from 2 years ago. I have not checked if the links are still valid, but at the time of the post all of the links and the phone number were valid.

It’s important for people to know. You can get slow insulin for $25 or so at Walmart. You can get the Walmart branded version of NovoLog for about $70ish dollars. Or you can get a generic version of NovoLog or Humalog for pretty cheap.

Or you can get actual REAL non-generic NovoLog for about $33. This is a real deal. At the time this came out, I did this as a test. I already had NovoLog covered so I got Levemir for $33 just to test if it worked. It did.

Comment from 2 years ago:

This is very easy to do. I tested it out by getting an insulin that was not on my insurance plan. It worked just as they described it! No gimmicks, no salary check, no insurance check.

Here is the reference thread: