Search for SPLE preparation and you will be offered a predicted score. The number looks precise, it is usually somewhere near 536, and it is not real. Understanding why is the difference between preparing against evidence and preparing against a comforting figure.
Every figure in this guide is checked against the official SCFHS source linked below - last verified 13 September 2026.
What SCFHS actually publishes
Three facts, and only three.
The exam is 200 four-option single-best-answer questions, delivered in two parts of 100 with 120 minutes for each part and a scheduled 30-minute break between them. Up to 10% of those questions are unscored pilot items.
Results are reported on a scaled range of 200 to 800.
The pass score is 536.
What SCFHS does not publish is any table, formula or worked example converting a raw mark into a scaled score. That absence is the whole story.
Why a scaled score is not a percentage
Licensing bodies scale scores so that results are comparable between sittings. Different forms of the same exam are not exactly equally difficult, so a fixed raw pass mark would mean the standard drifted with whichever questions happened to appear. Scaling corrects for that: the raw mark needed for a given scaled score moves between forms, while the standard the scaled score represents stays put.
Two consequences follow.
First, the same raw score can map to different scaled scores on different forms. There is no single conversion even in principle, let alone one that an outside party could derive.
Second, the unscored pilot items make outside reconstruction harder still. Up to twenty of the two hundred questions do not count, nobody is told which, and nobody outside the commission knows how they are distributed across the blueprint. Any external attempt to map raw to scaled is guessing at the denominator before it starts guessing at the curve.
So where do predicted scores come from?
They are invented - usually by taking a percentage on a practice bank and mapping it onto the 200-800 range with a linear rule somebody made up. The resulting number carries a precision it has not earned, and it fails in the direction that costs candidates money: a comfortable-looking 560 from a practice bank that skews easy tells you to book a test date you are not ready for.
You can test this yourself. Ask any product that gives you a predicted scaled score what conversion it uses and where the conversion is published. There is no published conversion to point at.
What to measure instead
Percentage correct, broken down by blueprint domain. It is less satisfying than a single number, and it is far more useful, because it answers the question a score cannot: what do I do on Monday.
The blueprint weights are Basic Biomedical Sciences 10%, Pharmaceutical Sciences 35%, Social, Behavioral and Administrative Sciences 20%, and Clinical Sciences 35%. A per-domain result tells you whether a 68% overall is a flat 68% or a strong clinical performance hiding a 45% in the Saudi-specific block. Those are different problems with different remedies and different timelines.
A workable way to read your own practice results:
- 80% and above across the domains - strong across the blueprint, with no domain dragging.
- 70 to 79% - on track; identify the weakest domain and close it before booking.
- 60 to 69% - borderline; target the two lowest domains rather than revising everything again.
- Below 60% - more revision needed across several domains, and a test date is premature.
Those bands describe performance on practice material. They are not a prediction of the scaled score, and they should never be presented as one - including here.
Three things worth doing with your practice data
Track the trend, not the last result. One mock is a noisy measurement. Three, spread over weeks, tell you whether revision is working.
Watch pacing separately from accuracy. At roughly 72 seconds per question, a domain where you are accurate but slow will read as a knowledge gap in the second part of the paper when it is really a time problem.
Check the second part against the first. Two 100-question parts with a break in between is a stamina exercise. If your accuracy falls between parts, the fix is sitting full two-part practice sessions, not more content revision.
And after you pass
The score keeps mattering. Candidates who pass may sit the exam twice more to improve the score, because it is used in residency selection. If a training programme is the goal, the question is not whether you cleared 536 but how far above it you are - which is another reason to prepare against a per-domain view rather than a pass-or-fail one.
Format, attempts and registration are covered in the SPLE exam guide, and the domains themselves in the SPLE blueprint explained.
Common questions
How many questions do I need right to score 536? Nobody outside SCFHS can tell you, because no raw-to-scaled conversion is published, and the answer would in any case differ between forms.
Is 536 out of 800 the same as 67%? No. The scale starts at 200, not 0, and it is not a percentage of anything. Dividing 536 by 800 is arithmetic on two numbers that have no such relationship.
Does negative marking affect the scaled score? There is no negative marking. A correct answer earns a mark and a wrong one costs nothing, so leaving questions blank only loses you marks.
When do results arrive? Two to six weeks after the test.
