Subway: the calories are on the board and people still guess low

Nothing on this page is a study of a sandwich. The evidence about Subway is evidence about menus, labels and what customers believe, and it is worth reading in that light. Putting calories on a menu board does work, and the effect is small. For an average 600 calorie meal the Cochrane review puts it at 11 calories less chosen, about 1.8 percent, rated high certainty, and 35 calories less actually eaten, about 5.9 percent, rated low certainty. Both numbers come out of the same review and the second one is the bigger of the two. Subway is where that small effect runs into a large reputation. Adults who had just bought a meal there guessed 20 percent lower than people buying at McDonald's, and teenagers guessed 25 percent lower, in a study that checked the actual receipts. The average adult fast food meal in that study held 836 calories and was underestimated by 175. And when New York made the chains post calories, the average purchase fell at McDonald's, Au Bon Pain and KFC, and rose at Subway, from 749 to 882 calories. Nobody has shown why.

The numbers, per 100 g

How much is actually in it

• Over five years New Zealand fast food servings got bigger and saltier, and Subway was one of four chains where that was worst. — 5,468 products across 10 chains and 12 food groups in New Zealand (n = 5,468 products) Evidence C source
  Mean serve size rose 9 g (+5%), energy density 54 kJ/100 g (+6%), energy per serve 178 kJ (+14%) and sodium per serve 55 mg (+12%); four food groups including Sandwiches and four chains including Subway showed large undesirable changes on three or more of five outcomes; the only significant reformulation change was a 7% fall in sodium density

What a real portion looks like

• Diners at Subway underestimated what they had eaten by a fifth more than diners at McDonald's. — 1,877 adults and 330 school-age children at dinnertime plus 1,178 adolescents after school or at lunchtime, across 89 restaurants of six chains in four New England cities in 2010 and 2011 (n = 3,385) Evidence C source
  Mean actual meal content was 836 kcal for adults, 756 for adolescents and 733 for children; underestimation averaged 175, 259 and 175 kcal respectively and grew as the meal grew; adults and adolescents at Subway estimated 20% and 25% lower than McDonald's diners (relative change 0.80, 95% CI 0.66 to 0.96 and 0.75, 0.57 to 0.99)
• Three customers in a hundred actually ordered from the low-fat range. — 927 respondents aged over 16 at Subway and McDonald's in Queensland, Australia, 73% response rate; the promoted ranges were Subway's Six grams of fat or less and McDonald's Heart Foundation Tick Approved (n = 927) Evidence C source
  Only 3% (24 of 910) of those ordering a main option bought a nutritionally promoted item; those who did were about 13 years older, 79% female and far more often working in a health-related profession (29% against 11%)

What it does to your health

• Calorie numbers on a menu shift an average order by about eleven calories. — 25 studies, 18 of them randomised; 16 conducted in real-world settings, 13 of those in restaurants or cafeterias; 24 in high-income countries (n = 9,850 participants in the selection analysis) Evidence A source
  Energy selected fell by a standardised mean difference of 0.06 (95% CI -0.08 to -0.03; 16 randomised studies, high-certainty evidence) and consumption may fall further (SMD -0.19, -0.33 to -0.05, low certainty); for a 600 kcal meal this is 11 kcal less selected (1.8%) and 35 kcal less consumed (5.9%)
• After New York put calories on menus, the average purchase fell at three chains and rose at Subway. — 7,309 adult customers in 2007 and 8,489 in 2009 at 168 randomly selected locations of the top 11 fast food chains, surveyed at lunchtime (n = 15,798) Evidence C source
  Across the full sample mean calories purchased did not change (828 against 846 kcal, p=0.22); McDonald's fell 829 to 785, Au Bon Pain 555 to 475 and KFC 927 to 868, while Subway rose 749 to 882 (p<0.001); the 15% of customers who said they used the information bought 106 fewer kcal

What people believe that the data does not support

• Showing the nutrition numbers hurt the Subway sandwich that broke the healthy image most of all. — participants randomly assigned to the information-present or information-absent condition and shown all four items (n = not reported in the abstract) Evidence B source
  Intentions fell for every item except the one that disconfirmed the health horn, and fell most sharply for the item that disconfirmed the health halo, the Subway Italian Spicy sandwich

Who this works differently for

• The small effect of calorie labels was the same for richer and poorer customers. — 17 eligible studies, 13 experimental and 3 pre-post implementation in the real world (n = 17 studies) Evidence A source
  No significant difference by socioeconomic position in either the experimental studies (chi-square 0.26, p=.610; higher SEP SMD 0.067, lower SEP SMD 0.115) or the real-world studies (chi-square 0.22, p=.636)
• When one chain added calorie labels, nothing measurable changed in what teenagers bought. — adults, adolescents and children buying meals at the labelling chain and at control chains; mean purchased meals were 804 kcal for adults, 741 for adolescents and 697 for children, most participants Black or Hispanic (n = not stated in this section) Evidence C source
  In adults labelling was associated with lower sugar content and lower percent calories from sugar but higher percent calories from saturated fat, and lower fibre; in adolescents there was no change in nutrient content or density at all; in children fibre content and density fell; sodium did not change in any group

The bottom line

The useful part is the order rather than the chain. In Queensland, 3 of every 100 people who bought a main item chose from a nutritionally promoted range, either Subway's six grams of fat or less menu or the ticked items at McDonald's. The people who did were about 13 years older than everyone else, 79 percent of them women, and nearly three times as likely to work in a health-related profession. A healthier menu that almost nobody orders from changes almost nothing. Two further things are worth carrying out of here. Servings at New Zealand fast food chains grew between 2012 and 2016, by 9 grams and 178 kilojoules per serve on average, with 55 milligrams more sodium. Subway was one of four chains where that drift was worst, on three or more of five measures. And when one chain added calorie labels, purchases by adolescents showed no change in nutrient content or nutrient density at all, while sodium moved in nobody, at any age. The small effect of labelling was also the same for richer and poorer customers, which is the answer to a common worry about who these rules help. Salt is the number this page owes you and does not show. The table above carries the four columns we publish for every food, and sodium is not one of them. Here it is from the same USDA rows. The eleven Subway subs we hold run from 268 to 575 milligrams of sodium per 100 grams, and a sub weighs a good deal more than 100 grams. The menu board will not tell you that one.

Sources

1. sub-dr-06 · repeated cross-sectional product survey
   Four of 12 food groups (Desserts, Pizza, Sandwiches, and Salads) and four of 10 fast food chains (Domino's, Hell Pizza, Pizza Hut, and Subway) displayed large, undesirable changes for three or more (of five) outcomes
   Nutr J, 2018 · checked 2026-09-22
2. sub-dr-03 · cross-sectional observational study
   Adults and adolescents eating at Subway estimated 20% and 25% lower calorie content than McDonald's diners (relative change 0.80, 95% confidence interval 0.66 to 0.96; 0.75, 0.57 to 0.99).
   BMJ, 2013 · checked 2026-09-22
3. sub-dr-05 · cross-sectional purchase survey
   Only 3 % (n 24/910) of respondents who ordered a main option had purchased a nutritionally promoted item.
   Public Health Nutr, 2012 · checked 2026-09-22
4. sub-dr-01 · Cochrane systematic review and meta-analysis
   These effect sizes suggest that, for an average meal of 600 kcal, adults exposed to calorie labelling would select 11 kcal less (equivalent to a 1.8% reduction), and consume 35 kcal less (equivalent to a 5.9% reduction).
   Cochrane Database Syst Rev, 2025 · checked 2026-09-22
5. sub-dr-04 · repeated cross-sectional survey, natural experiment
   Three major chains, which accounted for 42% of customers surveyed, showed significant reductions in mean energy per purchase (McDonald's 829 v 785 kcal, P = 0.02; Au Bon Pain 555 v 475 kcal, P<0.001; KFC 927 v 868 kcal, P<0.01), while mean energy content increased for one chain (Subway 749 v 882 kcal, P<0.001).
   BMJ, 2011 · checked 2026-09-22
6. sub-dr-07 · randomised factorial experiment on behavioural intentions
   In particular, behavioral intentions were most substantially weakened for the item that entailed a health halo disconfirmation (Italian Spicy sandwich).
   Appetite, 2020 · checked 2026-09-22
7. sub-dr-02 · systematic review and meta-analysis
   Meta-analyses of 13 experimental studies that predominantly examined hypothetical food and drink choices showed that energy labelling tended to be associated with a small reduction in energy content of selections that did not differ based on participant SEP (X2(1) = 0.26, p = .610).
   Int J Behav Nutr Phys Act, 2023 · checked 2026-09-22
8. sub-dr-08 · natural experiment with control chains
   Calorie labeling was not associated with changes in the nutrient content or nutrient density of purchases in adolescents.
   J Acad Nutr Diet, 2020 · checked 2026-09-22