Prepare for the BHS Level 3 Coach assessment by training one repeated decision: diagnose before you correct. Write session objectives in observable terms, run a three-point causal check (horse, rider, exercise) whenever something goes wrong, sequence flatwork using a recognised training scale, and progress jumping exercises by changing a single variable at a time. Rehearse explaining your reasoning out loud while a peer plays a rider. Administrative details such as entry requirements, assessment booking, and fees are set by the BHS and should be checked on their qualifications pages rather than inferred from study material.
Coaching a rider is not the same as riding a demonstration
At Level 3 the productive skill is producing change in a horse-rider pair through your words and planning, not proving your own riding. Distinguish instructing (directive, step-by-step commands) from coaching (observation-led, questioning, responsive), and decide deliberately which mode each part of a session needs.
Instructing has real value: a nervous rider on a fresh horse may need short, direct commands with no questions attached. Coaching becomes the stronger tool once the rider can execute the basics, because it forces the rider to feel and describe the change, which is what makes it stick between lessons. Compare the two directly when you practise: take one fault, such as a collapsing inside shoulder in trot, and script both a directive correction and a coaching sequence that asks the rider what they felt on the left rein versus the right.
Applying the distinction means planning moments, not just exercises. Mark two or three checkpoints in a session plan where you will stop the exercise, ask the rider a question, and check their answer against what the horse actually did. If a rider says the canter felt balanced but the horse's tempo visibly quickened through the corner, you have found the exact point where your coaching must gently replace agreement with observation. Practise that specific exchange out loud; it is a different skill from knowing the correction yourself.
Turning vague aims like 'improve the canter' into assessable objectives
Rewrite every session aim as something you can see or count: who does what, in which gait or over which poles, judged by which visible sign. An objective you cannot observe is an objective you cannot coach toward or honestly evaluate at the end.
Worked scenario one. A drafted plan reads: 'Objective: improve the rider's canter transitions.' That wording hides the usual mistake. The coach jumps straight to correcting the rider's aids, but the horse was hollow and hurried into canter, so the rider was bracing against a losing battle. The better decision is to first observe the horse's balance in trot and in transitions within canter, reset the objective as 'horse maintains a steady four-beat rhythm for a full 20-metre circle after the transition, rider's hands following', and only then address the rider's aid timing. Why it matters: correcting the rider without fixing the horse's balance teaches the rider that their aids do not work, which damages confidence and wastes the session.
Apply this rewriting habit to your whole plan. 'Improve the canter' becomes 'after a walk break, horse strikes into canter on the correct lead within two strides on a named bend, tempo unchanged for six strides'. Compare each rewritten objective against three checks: is the outcome visible, is it achievable in one session for this pair, and does the progression build toward it? If an objective fails any check, it is a theme rather than a plan, and themes cannot be coached in the moment.
Diagnosing whether the fault belongs to the rider, the horse, or the exercise
Before correcting anyone, run a quick three-point check: what is the horse doing, what is the rider doing, and what did my exercise actually ask for? Then change one variable, re-observe, and only escalate to a second change after seeing the result.
This habit prevents the common error of stacking corrections. A rider whose position collapses through turns may be weak, or may be compensating for a horse that falls in, or may simply have been asked to ride an exercise beyond the pair's current straightness. Practise the check as a spoken routine on paper: name the horse's response, name the rider's visible action, name what your exercise demanded, then write the single smallest change that tests your leading hypothesis. If the hypothesis was 'the exercise asks for more bend than the horse can hold', the test is to enlarge the circle, not to reposition the rider's leg.
Compare this with the unstructured alternative: firing three corrections at once ('heels down, look up, soften the hand') leaves neither you nor the rider able to say which change produced the improvement. Build a small library of paired scenarios in your notes, where each fault has at least two plausible causes and a different first test for each. Reviewing causes in pairs, rather than memorising one correction per fault, is what makes the diagnostic habit transfer to a live lesson.
Sequencing flatwork with a scale of training instead of a list of favourite exercises
Sequence a flatwork session from the bottom of a recognised training scale upward: confirm rhythm before asking for suppleness, suppleness before expecting a reliable contact, and so on. The table below turns each stage into something observable.
The value of a scale is that it tells you where to restart when something unravels. If a horse begins rushing and coming above the bit in a shoulder-in attempt, the fault usually lives lower down the scale than the movement itself, in rhythm or suppleness, so the fix is a return to simpler work that restores those, not more of the failing movement. Compare a session planned as a scale-climb with one planned as a list of exercises: the scale-climb version has a built-in diagnostic order, because each stage only earns the right to the next once its visible criteria appear.
Be careful to treat the scale as a teaching framework rather than a pass-fail checklist. A horse will not show each stage perfectly in sequence in every session; stiffness on one rein, weather, and the horse's fitness all shift where a pair sits on the scale on a given day. Your preparation task is to be fluent at describing, for any stage, what you would see in the horse, what you would see in the rider, and which earlier stage you would return to if it breaks down.
| Scale stage | What you observe in the horse | First retreat if it breaks down |
|---|---|---|
| Rhythm | Regular footfall at the same tempo on both reins and in transitions | Ride forward in walk and trot on large shapes; drop the demanding movement |
| Suppleness / relaxation | Even bend through turns, swinging back, no resistance on one rein | Return to large circles and changes of rein at the working gait |
| Contact | Steady, elastic connection; horse accepts the bit without leaning or hollowing | Recheck rhythm and suppleness in a longer, lower frame before re-closing the frame |
| Impulsion | Energy contained within the tempo; hind legs stepping actively under | Refresh with transitions within the gait before adding collection demands |
| Straightness | Hind feet tracking front feet; shoulders and quarters aligned on lines and curves | Return to shoulder-fore style positioning and counter-flexion on simple shapes |
| Collection | Shorter, more elevated strides carrying more weight behind, still forward-thinking | Reconfirm impulsion and straightness at working paces; shorten the session's ambition |
Choosing jumping progressions: changing one variable at a time
A jumping exercise changes in one dimension per step: pole to fence height, distance, approach, or related distance. When a pair struggles, reverse the last change rather than adding speed, height, or a second correction simultaneously.
Worked scenario two. A coach sets a trot-pole-to-cross-pole exercise; the rider launches forward over the cross pole, getting ahead of the movement, and the horse flats over the following single fence. The drafted plan's next step is raising the second fence to a vertical and shortening the related distance, on the theory that the pair needs to 'be tested'. That stacks three new demands onto a developing problem. The better decision is to keep the height identical, return to poles only until the rider's fold at the base is consistent, then reintroduce the cross pole at the same distance, changing nothing else. Why it matters: height is the variable riders fear most, and adding it while a positional fault is unresolved links jumping with anxiety and teaches the horse to anticipate a rushed approach.
Practise writing progressions as decision trees rather than linear plans. For each exercise, note the intended variable, the visible sign that the pair is ready for the next step, and the retreat position if it is not. Then rehearse the branch point aloud: 'Rider is ahead over the base; retreat one step, same height, simpler question.' This spoken rehearsal matters because in a live jumping session the temptation is to solve problems with the most dramatic available change, and a rehearsed retreat line is easier to say under that pressure than a fresh decision.
A film-and-annotate exercise to sharpen your observation between sessions
Film a short practice lesson you give to a peer or use publicly available schooling footage, then annotate it twice: once for what you said, once for what the horse and rider visibly did. Score it against the rubric below and repeat.
This is an observation drill, not a riding drill, so it carries no welfare risk: you are studying your own coaching words against visible outcomes. Split the screen or the transcript into two columns. In one, log every instruction you gave. In the other, log what changed in the horse or rider in the following thirty seconds. Then compare. The pattern to look for is instructions with no visible consequence, which usually means the instruction was not actionable, not checked for understanding, or aimed at the wrong layer of the problem.
Treat the scores as learning milestones, not as predictions about any assessment outcome. If your first annotation shows many unmatched instructions, that is simply your starting point. Repeat the exercise with a different rider type in the footage, for example a rider who talks through everything versus one who goes silent, because your coaching language needs to flex with the person receiving it. Keep your annotated notes together; they become your personal bank of coaching phrasings that demonstrably produced a change.
- Self-check rubric: count corrections given, and note how many were followed by a visible change within thirty seconds.
- Note how many corrections named an observable outcome ('tempo stays the same') rather than only a body position.
- Count questions asked, and whether the rider's answers were acknowledged or talked over.
- Check that only one variable changed between exercise repetitions.
- Record one moment you retreated a step successfully, or identify where you should have and did not.
An adaptable preparation sequence and honest readiness checks
Prepare in rotating cycles rather than a fixed calendar: plan a session on paper, deliver or simulate it aloud, annotate the result against your rubric, then rewrite the plan. Move on when the cycle produces fewer unmatched instructions across three consecutive cycles.
A workable sequence has four rotating strands: objective writing, fault-diagnosis scenarios in pairs, flatwork scale fluency, and jumping decision trees. Each cycle, pick one strand, produce something concrete, and test it in a simulation with a peer playing the rider, including at least one complication you did not script, such as a horse that spooks at a corner or a rider who cannot feel the wrong lead. Complications matter because a plan that only works in its intended version has not really been practised.
Readiness checks, framed as milestones for yourself: you can rewrite any vague aim into an observable objective in under a minute; given a described fault you can produce two plausible causes and a first test for each; you can recite, for each training scale stage, what you would see and where you would retreat; you can narrate a full jumping session aloud including two retreat points without hesitating; and your last three annotated simulations show more corrections with visible outcomes than without. If any check fails, that names your next cycle's strand. For administrative matters such as entry criteria and assessment booking, the BHS qualifications pages are the authority.
- Readiness check 1: any vague aim rewritten as an observable, single-session objective on demand.
- Readiness check 2: two causes and one first test produced for any described fault, on paper, unprompted.
- Readiness check 3: training scale stages described with visible criteria and retreat points, in your own words.
- Readiness check 4: a simulated jumping session narrated aloud with at least two decision-tree branches taken.
- Readiness check 5: three consecutive annotated simulations with improving correction-to-outcome ratios.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.