Scores Without Sanitation: Why Georgia's Prison Food-Safety Numbers Don't Reflect What Inmates Eat From

Brief written July 19, 2026 from GPS Intelligence System data.

Georgia’s county health departments inspect prison kitchens the same way they inspect a Waffle House: a sanitarian walks in, checks the sinks and the cooler temperatures, runs a thermometer through the sanitizing water, and leaves a score out of 100. Under Georgia law, a prison kitchen is a food service establishment like any other, subject to the same regulations and the same grading scale — A at 90 and above, B at 80 to 89, C at 70 to 79, and U, for unsatisfactory, below 70. On paper, the system works. Of the roughly 250 routine and follow-up inspections GPS examined at 36 state prison kitchens and staff dining halls from 2023 through early 2026, the overwhelming majority resulted in a score of 90 or above. More than one-third were perfect 100s. On paper, Georgia’s prison kitchens are among the safest places to eat in the state.

But the paper is the problem. A DPH score reflects a single announced walkthrough on a single day — a day when the kitchen staff knows the inspector is coming, when the broken dishwasher can be bypassed and the roaches can be chased into the wall, when the sewage that seeps through the floor drains the other 364 mornings of the year is mopped and masked. The men and women who work in those kitchens, and the thousands more who eat from their trays, describe a different reality. Georgia Prisoners’ Speak set out to find it.

Three Kitchens, One Pattern: The Failing Facilities

The most jarthing numbers in the DPH database are the ones that fall below 70. Three state prison kitchens have failed inspections since 2022: Johnson State Prison in Wrightsville (64 in December 2023), Pulaski State Prison in Hawkinsville (67 in January 2026), and Smith State Prison in Glennville (68 in May 2022). But the failures themselves are not the story. The story is what happens between the failures — the boom-and-bust cycle that shows the inspection regime catching the mess but doing nothing to stop it from reappearing.

Johnson State Prison. On December 11, 2023, a DPH inspector gave Johnson’s main kitchen a 64. The inspection report, obtained and analyzed by GPS in April 2026, described bulkn food items — oil, flour, and rice bran — with holes gnawed through the bags, visible rat droppings, and urine staining. The inspector wrote that “multiple rats and roaches” were present throughout the kitchen and that the problem was ongoing “with little to no change.” Five cooking ovens, one tilting skillet, one cooking kettle, one griddle, one freezer unit, and one bulk ice machine were all broken. Holes gaped in the floors, walls, and ceilings. Multiple cold-holding foods exceeded 41°F and had to be discarded. A follow-up inspection eight days later, on December 20, 2023, recorded a score of 67 — still failing — with cold-holding violations and equipment described as “in need of repair or replacement.” Five months later, in March 2024, the kitchen rebounded to 86, a B. By December 2024, it was back to 96, a clean A. But by March 2025, the score had slid to 80, a B, and in October 2025 it scraped back to 88, still a B. The pattern is a game of whack-a-mole: the inspector finds an infestation, the facility scrubs for the follow-up, the inspector leaves, and the rats return.

Pulaski State Prison. Pulaski’s fall was more dramatic and more public because it involved raw sewage. On January 29, 2026, the kitchen scored a 67, failing. GPS reported that the facility’s only designated handwashing sink was nonfunctional — the plumbing had been ripped from the wall with the pipe smashed inward. Sewage was backing up through floor drains, a violation the inspector marked as “repeat.” Employees switched between tasks without washing their hands. Nacho meat registered 65°F and sauce at 123°F — both hot-holding violations, also marked as repeats. The inspection was the culmination of a steady decline: 83 in February 2025, 73 in August 2025, a follow-up of 78 in September 2025, and finally the 67 in January 2026. A follow-up on February 6, 2026 brought the score to 96 — an A — but the underlying question remains: how did sewage back up through the floor drains again after being cited as a repeat violation months earlier?

Smith State Prison. Smith’s kitchen first failed in May 2022 with a 68. By June 2023, it had risen to 84, a B. It stayed in the low 80s — 82 in January 2024, 85 in June 2025 — before falling to a 72, a C, in February 2026. GPS’s review found that DPH inspectors noted rodent activity in every inspection at Smith from 2022 through 2025. The February 2026 inspection found roach activity in the bakery and tray-making station — marked as a repeat — along with broken handwashing sinks, broken plumbing in four sinks, a broken dishwasher faucet, clogged floor drains, and damaged walls with mildew on walls, floors, and ceilings (also marked as a repeat). The roaches were still there; the score just bounced because, on the day of the follow-up, the kitchen had been momentarily scrubbed.

The Clean Score Factory: High Grades Conceal What Inmates Eat From

If the failing facilities betray the limits of the inspection system, the high-scoring facilities expose a deeper falsehood. GPS has documented facility after facility where DPH scores are in the 90s or 100s while inmate accounts and photographic evidence show conditions that would earn a failing grade if the inspector came during actual meal service.

Carla Simmons, incarcerated at three Georgia women’s facilities since 2004, has written publicly about finding “roach legs in cornbread and rats climbing over the dry goods in the warehouse.” In Filter Magazine, an incarcerated writer using the name “Jimmy Iakovos” described a kitchen worker named “Bill” who was told to “shake the spoon” to short portions, menus designed for 150 that feed 200, and inmates whose “teeth are loose, bodies gray and bony” from the food.

GPS’s own investigation, published in May 2026 under the title “Dunked, Stacked, and Served,” assembled a body of evidence that connects high scores to contaminated trays. Inmate maintenance workers at Dooly State Prison reported thousands of roaches inside kitchen equipment. At a facility with a score in the 90s, photographs from Johnson State Prison showed serving trays caked with dark residue along the compartment seams, brown and black buildup clinging to the edges. Another resident, at Coastal State Prison, reported that meals were served on visibly contaminated trays. The Marshall Project, in a May 2026 investigation, quoted incarcerated individuals who said they were fed food contaminated with rats, insects, and mold, and that the state spent about 60 cents per meal.

The contradiction between the scores and the testimony is not a matter of bad data. It is a structural feature of how the inspections work. A DPH sanitarian walks through a kitchen that has been forewarned and watches a food tray being assembled. The inspector does not turn off the dishwashing machine to see if it can still wash. She does not interview the men cleaning the equipment about what they see when the machine breaks and trays sit in a barrel of chemicals overnight. She does not come at 5 a.m. when the roaches are active. The high scores are real in the sense that the inspector saw what she saw. But they are not real in the sense that they reflect what the man on the other end of the tray puts into his mouth.

The Inspection Regime That Cannot See: Scheduled Visits and Small-County Capture

The core weakness of the Georgia prison food-safety system is not that inspectors are corrupt or incompetent. It is that the inspection regime was designed for restaurants, not for institutions where the eaters cannot leave and the operators have every incentive to manage the inspection rather than the kitchen.

Georgia prison kitchen inspections are scheduled in advance. As The Marshall Project reported in March 2025, health departments must generally arrange their visits with prison officials due to security protocols, which gives staff time to prepare. Even when inspectors find violations, they are reluctant to shut down a prison kitchen because incarcerated people have no alternative food source — a blunt but practical reality that turns the enforcement lever into a toothless advisory. The DPH can give a failing grade, but it cannot close the kitchen.

There is no centralized database that tracks prison inspection trends across counties. Prison records are scattered in the statewide system alongside inspections for restaurants, hospitals, and schools. GPS located them by pulling the reports from individual facilities one by one. No annual report aggregates the data, no trend analysis flags facilities with repeated rodent or sewage violations, and no outside body is charged with connecting the scores to the plates.

In many counties, the prison is the largest employer, and the sanitarian who inspects the kitchen may know the warden socially, see the food-service director at church, or have professional relationships that make it hard to treat a state prison the same way she treats the local diner. GPS has documented that in small-county settings, the professional and social overlap between inspectors and facility staff raises a structural question — not of individual bad faith, but of whether a system designed for voluntary compliance in a competitive marketplace can ever be adequate when the consumer is a captive and the inspector is a neighbor. The GPS investigation “Dunked, Stacked, and Served” frames this as a regulatory-capture dynamic that hides the worst conditions from the scores. The state’s own system — dual inspections by GDC internal staff under Board of Corrections Rule 125-1-2-.10, which are not made public — does nothing to fill the gap.

The Kitchen the Scores Don’t Reach: SMU and the Gumm Settlement

One final note on what the inspection scores do not see. At Georgia Diagnostic and Classification State Prison (GDCP), the Special Management Unit (SMU) houses prisoners under extreme isolation. The food chain there historically did not pass through the same kitchen as the general population. In Gumm v. Ford, a federal case in the Middle District of Georgia, Timothy Gumm alleged that SMU inmates received food that was inedible, undercooked, cold, rotten, or expired, and that he lost 40 pounds. A class settlement approved in May 2019 required SMU prisoners to receive the same food as general population. That settlement suggests that whatever score the main kitchen at GDCP earned — and GDCP’s main kitchen scored a perfect 100 in March 2025 — it did not describe what reached the men in the SMU. The point is not to relitigate Gumm but to note that the SMU food chain is a gap in the inspection coverage, one that even a 100-point score cannot fill.

What a Score Does and Does Not Say

A DPH food-safety score is not a lie. It is a measurement of a narrow set of conditions recorded by a trained professional during a single announced visit. That measurement has value: it tells you that on March 26, 2026, at Telfair State Prison, the hot water was hot enough and the cooler was cold enough. What it does not tell you is whether, two weeks later, the dishwasher is broken and the trays are being washed by hand in a plastic barrel of chemicals, whether mice are running across the dry-goods shelves at night, whether the man who ate from that tray is vomiting in his cell. Those conditions are invisible to the scoring system because the scoring system was never asked to look.

That is the argument this page makes. Georgia’s prison food-safety numbers appear to show a system that is largely clean and well-run, with a few exceptions that get corrected quickly. In reality, the numbers show a system that is largely unobserved, punctuated by brief moments of accountability that fade as soon as the inspector walks out the door.

Sources

This analysis draws on DPH food-safety inspection records obtained from the Georgia Department of Public Health’s public database (ga.healthinspections.us) and from Open Records Requests to the DPH South Health District; GPS’s own investigative report “Food Safety Inspections in Georgia State Prisons” (April 2026), which first analyzed the inspection data and interview accounts; the GPS investigation “Dunked, Stacked, and Served”; the May 2026 Marshall Project article on Georgia prison food; Filter Magazine’s accounts from incarcerated workers; statements by Carla Simmons published in Truthout; federal court records in Gumm v. Ford; and inmate and family accounts collected and curated by Georgia Prisoners’ Speak.