A worksheet only records an exercise; the change happens somewhere before it. You can hand a child a beautifully designed thought record and still get nowhere, because the real work was in the conversation, and in the moment the child noticed their stomach tighten and found a word for it. The page just keeps a copy afterward.

That distinction matters when you're building a CBT practice for children and adolescents. Worksheets are the visible layer, and they're worth getting right (I wrote a separate guide to the ten that hold up). Underneath them sits the actual toolkit: a model the child can hold in their head, exercises that change behavior, and a few measures that tell you whether any of it is working. This post is about that layer.

Start with the model

Before any exercise, the child needs a frame to hang it on. In CBT that frame is the link between thoughts, feelings, and behavior: the CBT triangle, which families often look up as the schéma of how the three connect.

For an eight-year-old, draw it literally. Three circles, arrows between them, a real situation in the middle: "You walked into the party." Thought: "Nobody wants me here." Feeling: "Scared, 7 out of 10." Behavior: "Stood by the wall." Then walk the arrows backwards. If the thought had been "I'll find someone I know," how would the feeling change? What would the body do then? The point the child takes away is that the thought in the middle is the part they can change.

Teens can handle the model abstractly, but don't assume they want to. Draw the same triangle over a situation that actually stings, like a left-on-read text or a bad grade, and it lands harder than a clean diagram does. Build the model once, concretely, and every exercise after it has somewhere to live.

Measure something, from session one

Child CBT drifts without measurement. The kid says they feel "fine," the parent says nothing's changed, and six weeks in you can't tell whether you're helping. A handful of brief, validated questionnaires fix this, and they double as exercises in their own right, because completing them teaches self-observation.

SCARED (Screen for Child Anxiety Related Disorders) — child and parent versions, strong for sorting which anxiety is driving things.
RCADS (Revised Children's Anxiety and Depression Scale) — anxiety and low mood in one instrument, with subscales that track separately over time.
SMFQ (Short Mood and Feelings Questionnaire) — thirteen items, low enough burden to repeat every few weeks.
Spence Children's Anxiety Scale — useful when you want a finer anxiety profile.

Score them, plot the numbers, and show the child their own line going down. That graph is one of the most motivating tools in the room, and far more convincing than your telling them things are improving. Re-administer every three to four weeks rather than every session, so ordinary week-to-week noise doesn't read as deterioration.

Cognitive exercises that don't need much reading

The core cognitive move is to catch a thought, check it, and choose another. That is hard to do on paper with a child who reads reluctantly, so run it as an exercise instead.

Thought detective, out loud. Take a real worry the child names, treat it as a suspect, and gather evidence together verbally while you scribe. The investigation is the exercise. If you use a worksheet, it just stores the verdict.

The two-chair check. For older children and teens, physically move. One chair holds the anxious thought, the other the calmer alternative, and the child speaks from each. Moving between the chairs externalizes the two voices better than sitting still and reasoning ever does.

Worry postponement. Set a daily ten-minute "worry time." When a worry shows up outside it, the child notes it and parks it for later. By worry time, most worries have lost their charge, and the child learns from doing it that a thought isn't an emergency.

Behavioral exercises, where most of the change happens

Cognitive work gets the attention, but with children the behavioral exercises do most of the lifting.

Exposure, built as a ladder and then climbed. The hierarchy is the plan; the exercise is climbing one rung. Start low enough that the first step is almost easy, so an early success carries the child up the rest. Stay in each exposure until anxiety drops meaningfully, then repeat it before moving up. The mistake I see most is climbing too fast: a rung that spikes anxiety and gets abandoned teaches avoidance, the opposite of the point.

Behavioral experiments. Turn a belief into a test. "Nobody will talk to me if I sit with them" becomes a prediction, a plan, and a result. Teens respond well to the scientific framing, because it feels objective and rational.

Behavioral activation for low mood. Schedule one small pleasant or mastery activity a day, rate predicted enjoyment beforehand and actual enjoyment after. Depressed adolescents reliably under-predict, and seeing the gap between the two numbers is what shifts the belief that nothing will feel good.

Relaxation and grounding, practiced before they're needed. Diaphragmatic breathing, progressive muscle relaxation, and 5-4-3-2-1 grounding only work if they're rehearsed when the child is calm. A breathing exercise introduced for the first time mid-panic fails. Drill it as a dull weekly habit so it's automatic when it counts.

Tools that make the abstract concrete

A few props earn their place across almost every case:

The feelings thermometer — a 0-to-10 scale the child colors in, giving you a shared shorthand ("Where are you on the thermometer?") and the scaling that exposure work depends on.
A coping menu — preferred strategies sorted into body, mind, social, and creative, chosen by the child and kept somewhere reachable rather than in a therapy folder.
Emotion cards — for the younger end, or for any child who can point long before they can explain.

These overlap with the worksheet library, which is fine; a tool and its printable version are just two formats of the same thing. What matters here is using them live, in the room, as part of an exercise.

What shifts between kids and teens

The same toolkit, tuned differently.

Concreteness. Younger children need the model drawn, the metaphor literal, the exposure physical. Teens can work with the idea, once you've earned the right to be abstract with them.

Handing over control. With a younger child you build the ladder together. With a teen, hand over more: let them design the experiment, score their own questionnaire, set their own worry time. Ownership is itself therapeutic at that age.

Framing. A detective badge delights a nine-year-old and insults a fifteen-year-old. For teens, the framing that works is competence: here's how the technique works, here's the evidence, you run it.

Putting the toolkit together

Don't deploy all of this at once. A workable first arc: build the model in session one, start a brief questionnaire to set a baseline, teach one regulation exercise, and move into cognitive and exposure work as the case calls for it. The worksheets slot in as the record-keeping layer once the exercises are running.

A resource generator for child CBT can produce the printable side — thermometers, ladders, experiment logs, coping menus — in consistent, age-appropriate layouts, so prep time goes into planning the exercise instead of drawing it. The harder judgment, which exercise this child needs and how to take them through it, stays with you.

CBT Exercises and Tools for Children and Teens