Rheem Furnace Blinking Light Codes: Legacy LED Flashes Explained

Only one Rheem generation blinks: the legacy four-LED boards of the RGRM/RGTM era, with separate FLAME, STATUS, POWER and CFM lights. This guide covers that generation only — its STATUS table (1, 2, 3, 4 and 6 blinks, steady-on for normal, and 5 blinks explicitly 'Not applicable'), the FLAME and POWER lights that frame it, and the CFM light whose long airflow counts are the great miscount trap of these boards. It ends with the bridge every owner of a newer Rheem needs: models with a 7-segment display show two-digit codes instead, and no flash table — including this one — may be applied to them.

Which furnaces blink: the legacy four-LED generation, and only that generation

Start by identifying the reporter, because Rheem has used two incompatible ones. The blinking generation is the older Integrated Furnace Control with four separate LEDs labelled FLAME, STATUS, POWER and CFM — the RGRM/RGTM-era boards named in Rheem's manual 92-24161-84-02, and their Ruud equivalents. On these boards the STATUS LED genuinely blinks fault patterns, the FLAME LED reports flame quality, the POWER LED reports 24VAC power at the board, and the CFM LED blinks out the blower's target airflow. The newer generation — R95P, (-)801S/(-)801P, and the communicating (-)802V/(-)97V/R96V families — does not blink codes at all: its 7-segment display shows digits (0 in standby, H/C/F for operating modes, two-digit fault numbers shown one digit at a time), and reading those digits with a blink table manufactures faults that do not exist. The confusion is understandable, because a two-digit fault on a single display — first digit for a second, second digit for a second, a pause, repeat — can look like slow blinking to an owner counting flashes in a dark basement, and because the internet's Rheem 'flash tables' collapse both generations (and several other brands) into one list. This guide is deliberately narrow to prevent exactly that error: everything below describes the four-LED boards only. If your furnace has a digital display showing numbers or letters, close this page and use the two-digit chart for your model — the flame, limit and pressure guides on this site are written for that generation. If your furnace has the four labelled LEDs and no digit display, you are in the right place, and your board's own manual table — reproduced faithfully below, including the entry the internet most often gets wrong — is short enough to learn whole.

The four LEDs: what each light is actually reporting

The four LEDs divide the furnace's state between them, and reading the wrong one is the original sin of legacy-board diagnosis. The STATUS LED is the fault reporter: steady on means normal operation with no fault detected, and counted blinks (1, 2, 3, 4, 6) name the standing fault — its full table gets the next section. The FLAME LED reports flame quality rather than faults: off means no flame detected, a rapid blink means unexpected flame, a slow blink means marginal flame sense, and steady on means normal flame detected — a vocabulary that mirrors, in lights, what the display generation later split into codes 12 (low flame sense), 13 (flame lost) and 14 (unexpected flame). The POWER LED is the simplest and the most skipped: off means no 24VAC power to the control board, on means the control is powered — and a dark STATUS LED means nothing at all until the POWER LED has been checked, because a board without power cannot blink a fault code for the furnace that feeds it. The CFM LED is the trap, and it gets its own section below: it reports the blower's target airflow by blinking the number in hundreds of CFM, which produces long, confident blink counts — eleven and a half blinks is a normal reading on this light — that owners understandably mistake for an elaborate fault pattern. The reading discipline follows from the division: name the LED before counting it. A count taken from the CFM light and looked up in the STATUS table will always 'find' a fault, because the STATUS table's owner was never counting airflow. Photographing the whole board — all four lights labelled in frame — before counting anything is the cheapest accuracy upgrade available on this generation, and it is what a technician will ask for if you call mid-count.

The STATUS LED table, in full — including the blink that means nothing

Here is Rheem's STATUS table for the four-LED boards, complete, in the manual's own assignments. Steady ON: normal — no fault detected; the board is powered and satisfied, and a steady STATUS light beside a heating furnace is the end of the diagnostic conversation. 1 blink: ignition failure — 1-hour lockout; the furnace failed to establish and prove flame and has entered the same counted lockout the display generation later numbered 10 + 11. 2 blinks: pressure switch stuck open — the switch will not close when the inducer is on; draft was never proved and ignition was withheld (the display generation's 57 story). 3 blinks: limit switch open — main limit or roll-out control; the limit circuit has opened on over-temperature or roll-out (the display generation's 22/33 pair, undifferentiated on this board). 4 blinks: pressure switch stuck closed — the switch will not open; it was already made before the inducer ran, so the cycle never began (the display generation's 55 story). 5 blinks: Not applicable. Rheem's manual prints exactly that — the count is unassigned on these boards, it names no fault, and the section below deals with the famous claim that tries to give it one. 6 blinks: line and neutral reversed — the line-voltage polarity at the control is wrong (the display generation's code 26, which its manuals add 'or poor ground' to). Six entries and a deliberate blank: that is the whole table. Its shortness is its credibility — every count can be pointed to in Rheem's document for these boards — and it is also the tell for spotting padded internet versions, which stretch the list to eight or nine entries by importing other brands' assignments. If a chart gives your four-LED Rheem a meaning for 5 blinks, or any meaning beyond 6, it is not describing your board, whatever its headline says.

FLAME and POWER: the supporting lights that change the STATUS reading

The STATUS table does not stand alone, because the FLAME and POWER LEDs qualify everything it says — and in two common situations they overrule the reading an owner was about to act on. The POWER LED comes first in any diagnosis: if it is off, there is no 24VAC power to the control, and the correct checks are the household ones — the furnace switch, the breaker, the blower-door seating — not the fault table. A 'dead board with no blinks at all' is a power story until the POWER light proves otherwise, and technicians ask about that light before they ask about any count. The FLAME LED refines the STATUS 1-blink story in particular. Ignition failure (1 blink) says flame was not proved; the FLAME light says something about what the flame was doing meanwhile: a slow blink (marginal flame sense) during attempts points at the sense circuit — the rod and its connections — because flame existed but its proof was weak, the same branch the display generation numbers 12 on its way to 13 and 11. A rapid blink (unexpected flame) is the alarming one: flame reported when none should exist, the legacy rendering of the display generation's code 14, and it carries the same instruction — do not run the furnace; a gas valve, its wiring or the control's own sensing is indicted, and that is licensed work. Steady on (normal flame detected) beside a STATUS fault moves the story elsewhere entirely: the burners are lit and proved, so the standing blink belongs to whichever system the STATUS count names, not to ignition. Read as a trio, the lights give this old board a surprising amount of voice: POWER says whether the board can speak, STATUS names the fault family, FLAME grades the fire. Owners who report all three — 'POWER on, STATUS 3 blinks, FLAME steady' — hand a technician a board state that can be reasoned about over the phone, which on a generation without stored codes or a display is the closest thing to a fault history it can offer.

The CFM LED trap: airflow counts mistaken for fault codes

The CFM LED exists to tell a technician the blower's target airflow, and it does so by blinking the number in hundreds of CFM: one blink represents 100 CFM of target airflow, a half blink represents approximately 50 CFM, and when the sequence completes there is a noticeable pause of approximately one second before it repeats. Rheem's own worked example shows the scale of the trap: eleven and a half blinks, a pause, eleven and a half blinks again means a target airflow of approximately 1150 CFM — a completely healthy reading that an owner counting 'flashes' on the wrong light will experience as an alarming, unlisted 11-blink fault code, on a board whose real fault table stops at 6. Rheem hedges the number honestly, and the hedges matter: the value is target airflow, not measured airflow — manufacturing tolerances, air density, humidity and ambient temperature can move the actual delivered airflow significantly from the target — the resolution of the method is only 50 CFM (the manual puts the 1150 example's true target anywhere between 1125 and 1175 CFM), and the LED is intended for reference only. Three practical rules follow. First, a blink count above 6 is never a STATUS fault on these boards — it is the CFM light, or a miscount, full stop. Second, half blinks belong only to this light: if your count includes a fraction, you are reading airflow, not faults. Third, the CFM count is not an airflow measurement for diagnostic purposes — a furnace can blink a proud 1150 target while delivering far less through a loaded filter or collapsed duct, which is why limit faults (STATUS 3) are diagnosed on the air path itself, not on the target the board announces. The trap's persistence online — screenshots of CFM sequences captioned as fault codes — is one more reason the first step on this generation is always the same: name the LED, in the manual's labels, before counting anything it does.

The bridge: newer Rheems show digits, not blinks — do not cross the tables

Everything on this page stops at the edge of the four-LED generation, and the bridge to the newer one must be crossed consciously, because the two systems share a brand, a cabinet and almost no vocabulary. A newer Rheem — single 7-segment display on mainstream models, dual display on communicating ones — shows standby as 0, operation as letters (H for gas heat, C for compressor, F for continuous fan, lower-case on communicating models for low-stage calls), and faults as two-digit numbers presented one digit at a time with a half-second pause, each fault shown at least twice, paired faults alternating with the display's top segment lit between them. Nothing in that presentation is a blink count, and applying this page's table to it produces confident nonsense: a display cycling 1-then-2 for code 12 (low flame sense) is not '1 blink' (ignition lockout) or '2 blinks' (pressure switch) — it is a different code on a different reporter that happens to involve the same digits. The reverse error exists too: owners of four-LED boards search their blink counts against the two-digit chart and find meanings (12, 13, 14…) their board cannot express. Rheem's own documentation keeps the generations in separate manuals for a reason, and your furnace's model number and door chart settle which one you own in seconds: four labelled LEDs, this page; a digit display, the two-digit chart and the flame/limit/pressure guides on this site. One mapping is worth carrying across the bridge, carefully, as orientation rather than equivalence: the legacy subjects survive the redesign — ignition failure, pressure proving in both directions, limit opening, reversed polarity all exist on the display generation as codes 11, 55/57, 22/33 and 26 — so the fault families in this guide will feel familiar there, under numbers that must be read as numbers. What does not survive is any specific count-to-meaning pairing. Honour the reporter's identity, and both generations read cleanly.

Refuted: '5 flashes = twinning' and the padded flash tables

One circulating claim about these boards deserves a formal retirement, because it is specific, confident and contradicted by Rheem's own manual: the claim that 5 flashes on an older Rheem indicates a twinning fault. Rheem's STATUS table for the RGRM/RGTM boards states that 5 blinks is 'Not applicable' — the count is unassigned, full stop. No Rheem manual checked for this site documents a 5-blink twinning meaning on any board, and the claim's likely origin is visible in the display generation's chart: twinning genuinely is a Rheem furnace fault there — code 20, declared when a twinned pair (two furnaces on a common duct system and thermostat, linked by the TWIN terminal) is detected and then loses its link in the same power cycle — but code 20 is a displayed number on a 7-segment control, not a blink count on a four-LED one, and a single, non-twinned furnace should never show either. The padded tables that carry the 5-flash claim commit the family's other sins alongside it: Carrier-style flash assignments relabelled as Rheem, the genuine six-entry STATUS list stretched toward nine, and genuine display codes redescribed as flash counts until no generation can be recognised in the result. The defence is the one this site applies everywhere: a real code can be pointed to, in a named manufacturer document, with its definition in the manufacturer's words — as every entry in this guide can be (Rheem manual 92-24161-84-02 for the table itself). Where a count has no such home, the honest publication is the one Rheem printed: not applicable. If your four-LED board appears to blink 5, recount with the LED named and a full pause watched — the likeliest truths are a CFM sequence misread, a 4 or 6 miscounted mid-pattern, or a board whose own door chart (which outranks every generic list, including this one) assigns its counts differently.

Reading a legacy board well: counting discipline and when to call

A short table rewards careful reading, and three habits make a four-LED report trustworthy. First, count in full repetitions: watch at least three complete STATUS sequences, allowing for the pause structure, and video the board if the count is contested — joining a pattern mid-cycle manufactures codes that do not exist, exactly as it does on the display generation. Second, record the trio, not the count: POWER (on or off), STATUS (steady or its count), FLAME (off, slow, rapid or steady), plus what the furnace was doing — inducer running or silent, burners lit or not, blower moving air or still. On a board with no stored history and no fault buffer, that contemporaneous note is the entire diagnostic record, and a technician reconstructs the event from it. Third, let each count route you the way the display-generation guides route their codes: 2 and 4 are the pressure proof failing in opposite directions (clear the outdoor terminations for the stuck-open story; leave a stuck-closed switch and any suspected bypass to a technician); 3 is the limit circuit (replace the filter, open the air path, prove the blower — and treat any roll-out involvement as the manual-reset, technician-first event the limit guide describes); 6 is polarity (licensed electrical work — do not swap wires by trial, because polarity faults also corrupt flame sensing); rapid FLAME is unexpected flame (furnace off, licensed call). The call itself is due whenever: the count is 1 repeatedly after one power retest with gas proven; FLAME blinks rapidly at any point; STATUS 3 returns after the filter and registers are honestly sorted; POWER stays off with breaker, switch and door all proven; or the board's lights behave in a way no entry in its own short table describes. These boards served for years with a six-line fault vocabulary; read in their own dialect, they still say everything a homeowner needs to decide the only two questions that matter — what family is it, and is it mine to fix or a technician's to prove.

Key takeaways

• Name the LED before you count it: only STATUS blinks faults (1, 2, 3, 4, 6 — and steady-on for normal). A count above 6, or any count with a half blink, is the CFM airflow light, not a fault.
• 5 blinks means 'Not applicable' in Rheem's own RGRM/RGTM table. Any chart assigning it a meaning — including 'twinning fault', which is display-code 20 on the newer generation — is not describing your board.
• If your Rheem shows digits or letters on a 7-segment display, it does not use blink codes at all: read it with the two-digit chart, and never apply a flash table to it.

Resetting the Rheem furnace

Rheem fault codes are shown as two-digit codes on the control's 7-segment display and clear once the fault condition is gone. Lockouts follow Rheem's documented rules: soft lockouts (the one-hour lockout, code 10, and the five-minute lockout when the low pressure switch fails to close after 60 seconds of inducer operation) auto-reset when their timer expires, and any lockout can be cleared immediately — once the original fault is cleared — by cycling power to the unit off and back on, or, for soft lockouts only, by cycling the heat call off for more than 2 seconds but less than 20 seconds. Hard lockouts are the exception: code 93 cannot be reset by cycling the heat call, and a dead-blower hard lockout (code 61, declared four times in one heat call) needs the repair plus a manual reset of power. Clearing a lockout without fixing its cause just restarts the count to the next one.

Frequently asked questions

Related guides

Facts on this page were checked against manufacturer documentation: pts.myrheem.com · www.manualslib.com · pts.myrheem.com · pts.myrheem.com · www.manualslib.com · my.ruud.com:443