Research transparency: This guide interprets U.S. Department of Energy building-envelope guidance as a household observation exercise. It is not an energy audit, and it currently contains no affiliate links.

Windows receive blame for almost every uncomfortable edge of a room. The chair feels cold, the curtain moves, water appears at the bottom of the glass, or afternoon sun turns a bedroom into a greenhouse. Replacement may be justified, but the symptom alone does not identify the failure. The leak may be around the frame rather than through the unit. The sensation may be radiant cooling from cold glass without moving air. Condensation may point toward indoor humidity, surface temperature or both.

Before asking a salesperson to diagnose the house through a product catalog, spend one week observing the window that bothers you most. The aim is modest: learn when discomfort appears, where it is strongest and what changes it. Seven nights will not produce laboratory data. They can produce a far better brief for an energy auditor, installer or landlord.

Night one: draw the comfort map

Choose one room and one normal evening. Do not begin by sealing, draping or moving furniture. Sit where the household usually sits and mark three positions: beside the window, at the center of the room and near the interior wall. At each position, note the time, outdoor weather, thermostat setting and what your body notices.

Cold glass can make a nearby person lose radiant heat even when room air is not moving much. That can feel like a draft. A real air leak is more directional: it often appears at a sash, trim joint, lock, outlet or baseboard. Use the back of a dry hand to compare locations. For a more formal diagnosis, the Department of Energy recommends a professional energy assessment; blower-door testing and infrared imaging can show leakage that a casual check cannot.

Night two: find the path, not just the pane

Inspect the perimeter slowly. Is the moving air at the meeting rail, where two window sections close? At the operable sash? Between trim and wall? At the floor directly below? A window opening is a system of glass, frame, hardware, seals, rough opening and adjacent wall. Replacing the glass unit cannot repair every part of that system.

Do not use flame as a detector. A thin strip of tissue held safely near the closed frame can show movement without creating a fire risk. Mark the location with removable tape and photograph it. Repeat on a windy and a calm night. If movement only appears during strong wind from one direction, write that down; exposure matters.

Night three: read water as evidence

Condensation belongs in its own column. Record whether moisture appears on the room side of the glass, between panes or around the wall. Interior surface condensation can occur when moist indoor air meets a cold surface. Moisture trapped between sealed panes suggests a different issue. Staining or softness around trim may indicate water entry and deserves prompt professional investigation rather than a cosmetic cover-up.

Also note the household conditions: showering, cooking, drying laundry, running a humidifier and closing bedroom doors all affect moisture. The Department of Energy notes that excess condensation can damage window components and surrounding surfaces and encourage mold and mildew. The useful question is not simply “Is the window wet?” but “Which surface is wet, under what conditions, and how often?”

Night four: move the fabric

Now change one reversible layer. Close the existing curtains or shade and compare the same three sitting positions. Do not assume every covering insulates equally. Fit and edge gaps matter, and some products are better at privacy or glare control than heat flow. The Department of Energy's window-attachment research identifies cellular shades and insulating panels among the more effective options in heating-dominated conditions, but climate, orientation and product construction change the result.

If the closed covering improves comfort but creates heavy condensation behind it, record that too. The fabric may be reducing heat reaching the glass while trapping moist air. A successful intervention has to improve the room without hiding a moisture problem.

Night five: watch the sun leave the room

For a west- or south-facing window, begin before sunset. Record where direct sun lands, when the room peaks in temperature and how long it remains warm. Solar gain that is welcome in winter may be difficult in summer. Exterior shade, interior coverings, glazing performance and tree cover act differently; a replacement specification should respond to the climate and orientation rather than applying one glass package to every elevation.

When comparing new windows, learn two numbers. U-factor describes heat transfer; lower generally means better insulating performance. Solar heat gain coefficient, or SHGC, describes how much solar heat passes through; the useful level depends on climate and orientation. Visible transmittance matters too, because reducing solar gain should not automatically turn a bright room dim.

Night six: separate maintenance from replacement

Operate the window. Does it close squarely? Does the lock pull the sash tight? Is weatherstripping missing, compressed or painted over? Is caulk visibly separated at an appropriate exterior joint? Routine repair can restore performance when the main unit is sound. The exact repair depends on window type and water-management details, so avoid indiscriminately sealing drainage paths or applying expanding foam where it can distort a frame.

DOE building-science guidance notes that gaps between the window unit and rough opening can be major leakage sites and that low-expansion products intended for windows and doors are used in appropriate assemblies. This is different from smearing caulk over every visible seam. If trim removal, flashing or concealed moisture is involved, bring in a qualified professional.

Night seven: write the brief

Finish the week with a one-page diagnosis, not a shopping list. Rank the symptoms by importance and frequency. Include room, orientation, times, weather, exact leak locations, condensation observations, hardware condition and the effect of coverings. Add photographs. Then state what outcome matters: a warmer reading chair, less street noise, easier operation, lower summer gain, reduced condensation risk or a combination.

This brief changes the conversation. An auditor can test a specific concern. A contractor can explain which part of a proposed assembly addresses it. A landlord receives dates and locations instead of a vague report. And if replacement is still the right answer, the household can compare proposals against the problem rather than against a list of premium features.

What you observeWhat it may justify next
Directional air movement at an operable jointHardware adjustment or weatherstripping assessment
Air movement between trim and wallProfessional review of the rough-opening air seal
Cold sensation without detectable airflowGlazing and whole-window performance comparison
Interior condensation during high humidityHumidity, ventilation and surface-temperature review
Moisture between panesInspection of the insulated glass unit or warranty
Strong afternoon overheatingOrientation-specific shading and SHGC strategy
The best window proposal begins with a map of discomfort, not a promise that every old window is the problem.

Your seven-line diary

  • Time, outdoor conditions and thermostat setting.
  • Where discomfort is felt and whether air movement is directional.
  • Exact frame, sash, trim or wall location.
  • Condensation surface and household moisture activity.
  • Effect of closing the existing covering.
  • Effect of sun and orientation during the day.
  • The outcome the room needs, stated without naming a product.

Our selection frame

If replacement is supported by the evidence, we would compare whole-window U-factor, climate-appropriate SHGC, air-leakage information, condensation resistance where available, installation details, warranty responsibility, operability and the contractor's plan for the rough opening. Glass alone does not define the result.

Research sources