How the two differ in kind
An ice dam is a heat problem that presents as a water problem. Heat escaping from the building warms the upper roof deck, snow on it melts, the meltwater runs down and reaches the eave — which overhangs unheated space and is therefore cold — and there it refreezes. The ridge of ice that builds up dams the next run of meltwater, which ponds behind it, sits against the roof covering, and works its way under the shingles by capillary action and hydraulic pressure. Shingles shed running water; they do not resist standing water, and that distinction is the whole failure mechanism.
The two interventions here attack opposite ends of that chain. VENTILATION, combined with insulation and air sealing, keeps the roof deck cold enough that the snow above it does not melt in the first place — no meltwater, no dam. It is the cure, it improves the roof's summer behaviour and its moisture behaviour year-round, and it is difficult or impossible to retrofit on some roof forms.
The MEMBRANE assumes a dam will form regardless and makes the eave watertight underneath the covering. It is a self-adhering waterproof sheet that seals around fasteners, run from the eave up past the point where the warm interior below meets the cold overhang beyond — which is why codes express its extent as a distance measured up the slope past the inside face of the exterior wall, not as a fixed number of courses. It cures nothing and it prevents the damage that actually costs money.
The reason the honest recommendation is usually BOTH is that each fails where the other works. A perfectly ventilated roof still gets dams in a freeze-thaw cycle driven by sun rather than by escaping heat. And a membraned roof with a warm attic still forms dams, still loads the gutters and eaves with ice, and still soaks the insulation via whatever paths the escaping air is already using.
The factors that actually differ
Show
| Ice and water membrane at the eaves | Attic ventilation | |
|---|---|---|
| What it addresses | The consequence. Water that is already ponding cannot reach the deck or the structure. | The cause. A cold deck does not melt the snow above it, so no dam forms. |
| When it can be installed | When the roof covering is off — a re-roof, a new build, or a repair. It goes under the covering, not over it. | At any time in principle, though intakes at the eaves usually mean working from outside and baffles mean working inside. |
| Extent required | Measured up the slope from the eave to past the inside face of the exterior wall, so a low-pitch roof or a deep overhang needs more than one course. | Expressed as a ratio of net free vent area to the ceiling area it serves, split between low intake and high exhaust. |
| Depends on air sealing | No. It works whether or not the ceiling leaks. | Completely. Ventilating an attic fed by warm leaking air moves that air around rather than removing the heat source. |
| Other benefits | None beyond waterproofing — though it is also the right underlayment at valleys, around penetrations and on low slopes. | Substantial: lower summer deck temperatures, moisture control, longer covering life, less condensation on the underside of the deck. |
| Roof forms that defeat it | Almost none; it is applied to the deck whatever the shape. | Cathedral ceilings, complex hips and dormers, and roofs with no usable eave — often solved by an unvented assembly with insulation above or against the deck instead. |
| Failure mode | Under-extent. Membrane that stops short of the warm-to-cold transition puts the dam's water on unprotected deck. | Unbalanced or blocked. Insulation stuffed into the eave, no baffles, or exhaust with no matching intake — the flow stops and the figures on paper mean nothing. |
| Ice on the eaves and gutters | Unchanged. The dam still forms and still loads the gutter; the water simply does not get in. | Reduced or eliminated, which also removes the falling-ice hazard and the gutter damage. |
| Effect on drying | It is a vapour barrier on the cold side of the deck. On a ventilated roof that is fine; on an unvented assembly it is a decision the whole build-up has to accommodate. | It is the drying mechanism. A ventilated deck dries upward into the airflow, which is most of why ventilation exists at all. |
| Where the cost lands | Per square metre of eave, once, during a re-roof. Trivial if the covering is off and disproportionate if it is not. | Per vent plus labour, plus baffles and air sealing — and the air sealing is usually the largest and most worthwhile part of the bill. |
Which one, and when
Choose ice and water membrane at the eaves when…
- The roof is being re-covered — this is the one moment the membrane costs almost nothing extra to add.
- The roof form makes ventilation impractical: a cathedral ceiling, a complex hip, no usable eave.
- The climate has a snow load and a freeze-thaw cycle, where a dam is a question of when rather than whether.
- The eave detail is low-pitch, deep-overhang, or a valley — the geometries that pond water regardless of deck temperature.
Choose attic ventilation when…
- The attic is accessible and the ceiling below it can be air sealed at the same time.
- There is evidence of a warm attic: melt patterns on the snow, frost on the underside of the deck, ice at the eaves every winter.
- Summer heat, shingle life or attic moisture are problems in their own right.
- The roof covering is not due for replacement, so the membrane cannot be installed without stripping it.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- Will ventilation alone stop ice dams?
- Often, but not always, and only if the air sealing is done with it. Ventilation removes heat that has already entered the attic; it does nothing about heat that keeps arriving. In a typical house the dominant path is air leakage — around recessed lights, the loft hatch, plumbing stacks, wiring penetrations, the tops of partition walls — and until those are sealed, ventilating harder simply draws more warm air out of the house and into the roof. Even a well-sealed and well-ventilated roof can form dams when the sun warms a south-facing slope enough to melt snow while the eave stays in shadow below freezing, which is a weather mechanism rather than a building defect. That residual case is exactly what the membrane is for.
- How far up the roof does the membrane need to go?
- Past the point where the warm building below ends and the cold overhang begins, measured along the slope — that transition is where the meltwater refreezes and where the dam's upper edge sits. Codes in cold regions state it as a minimum distance measured up the slope from the eave edge to a point beyond the inside face of the exterior wall. Two geometries make that more than a single roll's width: a shallow pitch, because a given rise takes much more slope length to reach; and a deep overhang, because the transition point starts further up. Both are routinely under-estimated by counting courses rather than measuring, and membrane that stops short leaves unprotected deck exactly where the water stands.
- Do I need both?
- In a genuine snow-and-freeze climate, yes, and they are not competing for the same budget. Ventilation and air sealing reduce how often and how severely dams form, which protects the gutters, the eaves and everyone walking underneath, and pays for itself in summer heat and moisture control as well. The membrane accepts that some dams will form anyway and makes those events harmless to the structure. Where a choice genuinely has to be made, the trigger is usually timing rather than preference: if the roof is being stripped, the membrane is nearly free to add and would be expensive to add later; if it is not, ventilation and air sealing are what can be done this year.
- How much ventilation does an attic need?
- It is specified as a ratio of net free vent area to the area of the ceiling below, with the total split between low intake at the eaves and high exhaust at or near the ridge — commonly close to an even split, with intake never less than exhaust. Two details decide whether the built result matches the calculation. NET FREE AREA is not the size of the hole: louvres, insect mesh and weather hoods reduce it substantially, and the manufacturer's stated net free area is the number to use rather than the physical dimensions. And balance matters more than total: exhaust with insufficient intake pulls its make-up air from the house through every ceiling leak, which is the opposite of the intended effect and actively worsens ice dams.
- What about a heating cable in the gutter?
- It is a symptomatic treatment and it has a legitimate narrow use. A cable melts a channel through the dam so meltwater has a route off the roof, which can rescue a problem eave on a roof where neither ventilation nor a membrane is practical and re-roofing is years away. What it is not is a solution: it runs on electricity all winter, it addresses one location while the roof carries on forming ice everywhere else, and it fails silently in exactly the weather when it is needed. Treat it as a last resort on a specific troublesome detail, and not as a reason to skip air sealing or to omit membrane during a re-roof.
- My cathedral ceiling has no attic — what then?
- The conventional pairing does not apply and the roof is usually built as a different assembly. Ventilating a cathedral roof means maintaining a continuous air channel from eave to ridge above the insulation in every rafter bay, which is easy to draw and easy to block with insulation at the eave, and impossible across hips, valleys and dormers where bays are interrupted. The alternative is an unvented assembly with sufficient insulation above or in continuous contact with the deck, so the deck stays close to indoor temperature and never develops the warm-above, cold-below split that drives a dam. Both routes rely on an airtight ceiling plane, and both are design decisions rather than retrofits — which makes the membrane, on this roof form, more important rather than less.
- Can the membrane itself cause a problem?
- It can, on the wrong assembly, because a self-adhering membrane is close to vapour-impermeable and it sits on the cold side of the deck. On a conventionally ventilated roof that is not an issue: the deck dries upward into the ventilated space beneath it. On an unvented assembly, or where the membrane has been run over the whole roof rather than at the eaves and valleys, moisture that gets into the deck has one fewer route out, and the deck can stay wet long enough to matter. The rule that keeps it safe is the ordinary one for any assembly: know which side dries, and do not put a vapour-impermeable layer on that side. Where a fully membraned deck is genuinely wanted, the assembly is designed around it rather than the membrane being added to a design that assumed it was not there.
- Do gutters cause ice dams?
- No — they collect the evidence and take the damage. A dam forms at the eave because that is where the warm roof ends and the cold overhang begins, and it would form there on a roof with no gutter at all. What the gutter does is fill with ice, which gives the dam something to build against, adds considerable weight to a fascia that was never designed for it, and pulls the gutter off the building. Removing the gutters therefore does not stop the dams and creates a different problem at ground level. The useful interventions are the ones above: stop the heat getting into the roof, and make the eave watertight so that when a dam does form, the water behind it stays outside the structure.