A frequent question we get is how well Nima will be able to detect cross-contamination. People are concerned about a slice of bread that was cut with a knife used to cut a slice of rye bread or the impact of stray crouton dust.

To understand how real the threat of cross-contamination is and how much we might or might not capture with our device, we designed an experiment where we ordered food from 18 restaurants, ranging from fast food ($) to high-end restaurants ($$$$$). For each order, we indicated that the recipient either had celiac disease or had a severe gluten intolerance. We had this food either delivered or picked up and brought to the office. For items picked up in person, a confirmation that the food was gluten-free was verbally requested. In some cases, restaurant staff went above and beyond to explain separate fryers or to inform us that our original request was not gluten-free and guided us to more appropriate options (much appreciated, restaurants!).

Once back in our office, we took a spot sample from each part of the dish. So for a pizza, we sampled the toppings and then the crust. In the case of a meal, we might take a small sample of the main course and the sides separately.

We then emulsified the remainder of the meal and took a sample of the entire plate to determine the level of gluten across the entire dish.

We followed lab protocol, documenting each item in a spreadsheet, and with photos, bagging each item carefully for processing.

Our findings are largely inconclusive. In all cases, the food qualifies as gluten-free, that is, everything was found to be under 20 parts per million (ppm) of gluten.


  • In five cases, we did detect gluten between 10 and 20 ppm.
  • In another five of 18 cases, we detected gluten between 5 – 10 ppm.
  • In three of 18 cases, the emulsified dish detected a higher level of gluten than we found from spot detection.
  • In three of 18 cases, the spot detection found a higher level of gluten than in the emulsified dishes.
  • Only one case had showed no gluten at the spot detection level but did find gluten at the emulsified level.

cross contamination gluten test results

One of the reasons we looked at a range of prices was that people often assume that “fancier” restaurants are safer.^^ This wasn’t necessarily true, as only one of our five fast food restaurants showed traces of gluten — and was still below 10 ppm. The mid-tier seemed to be the biggest problem area, with three of seven showing gluten between 5 and 20 ppm. Of the most expensive tiers, two of five restaurants had gluten between 5 and 20 ppm. Interestingly, in both cases, we found gluten in the spot check, but not in the emulsified food.

What conclusions can we draw from this? The main one is that you have to be vigilant no matter what. You may find that spot testing food finds gluten that you wouldn’t encounter over the whole dish. Nima will definitely help you to find some hot spots, but you’ll still need to proceed with caution in those times where you get a smiley face. It’s the gluten found that should make you send that dish back!

^^To be truthful, it was also to make our staff cry that they weren’t able to eat beautiful dishes like these.

testing sushi for gluten

testing tuna nicoise for gluten