All AppsNeuron Trainer HealthNeuron Trainer TradesNeuron Trainer TechNeuron Trainer NextNeuron Trainer MotionStudy GuidesTechnologyRoadmapPricingReviewsAboutFAQFor OrganizationsSupportGet the Apps →
Free PL-300 practice · Tech

PL-300 practice test: 4 free Power BI questions, DAX and Power Query included

Microsoft’s PL-300 gives you 100 minutes and you pass at 700 or above on a scale to 1,000. It covers preparing, modeling and visualizing data at 25–30% each, plus managing and securing Power BI at 15–20%. There are no enforced prerequisites, but it expects working fluency in Power Query and DAX, and the certification renews every 12 months.

Formally the Microsoft Certified: Power BI Data Analyst Associate (Exam PL-300). Awarding body: Microsoft. Neuron Trainer is an independent study app and is not affiliated with, endorsed by or sponsored by the organization that owns this exam.

PL-300 at a glance, with the source for every line

Every row below is quoted from a document published by Microsoft and carries the date we checked it. Where Microsoft publishes nothing, the row says so instead of quoting a number from somewhere else.

Time limit
100 minutes.Source: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
Passing score
700 or greater, on a scale to 1,000.Source: Microsoft Learn, Exam PL-300 study guide, verified 7 August 2026
How long it lasts
Renewal frequency is 12 months. Microsoft Associate certifications expire annually and renew free through an unproctored online assessment on Microsoft Learn.Source: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
Prerequisites
None enforced. Microsoft expects proficiency with Power Query and DAX.Source: Microsoft Learn, Exam PL-300 study guide, verified 7 August 2026
Format
Proctored, and it may include interactive components.Source: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
Skills measured as of
20 April 2026.Source: Microsoft Learn, Exam PL-300 study guide, verified 7 August 2026
Cost
Microsoft publishes no single figure. Its PL-300 page states: “Price based on the country or region in which the exam is proctored.” You see yours at booking.Source: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
If you fail
You can retake it 24 hours after the first attempt. Microsoft states that for subsequent retakes the amount of time varies.Source: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
AZ-900, for comparison
AZ-900 is a Microsoft Fundamentals certification. Microsoft states that Fundamentals certifications do not expire, while its Associate certifications renew every 12 months.Source: Microsoft Learn, certification renewal, verified 7 August 2026
Question count
Not published by Microsoft. Microsoft does not publish a question count for PL-300. It publishes the 100-minute duration and the skills measured, and item counts vary between forms.Checked against: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026
Pass rate
Not published by Microsoft. Microsoft does not publish a pass rate for any of its certification exams.Checked against: Microsoft Learn, Power BI Data Analyst Associate, verified 7 August 2026

Always verify: Microsoft updates the PL-300 skills measured periodically, prices the exam by the country it is proctored in, and expires this certification annually. Confirm details at learn.microsoft.com before you book.

What Microsoft publishes as the official PL-300 blueprint

These weights are Microsoft's, quoted from the source below. They are the only weights on this page that describe the real exam.

Published by Microsoft and verified on 7 August 2026.
AreaWeight Microsoft publishes
Prepare the data25–30%
Model the data25–30%
Visualize and analyze the data25–30%
Manage and secure Power BI15–20%

Scroll the table sideways to see every column.

Source: Microsoft Learn, Exam PL-300 study guide, verified 7 August 2026.

What our PL-300 question bank covers

What this app's question bank covers — our own category shares across 651 questions, not Microsoft's published blueprint.
CategoryQuestions in our bankShare of our bank
Prepare the Data
Get data from sources, profile and clean data, transform and load using Power Query - 25–30% of exam
18228%
Model the Data
Design a data model, relationships, DAX measures and calculated columns, time intelligence, row-level security - 25–30% of exam
18228%
Visualize & Analyze
Reports and dashboards, visualization types, drillthrough, tooltips, accessibility, advanced analytics - 25–30% of exam
17627%
Manage & Secure Power BI
Workspaces, apps, and distribution; dashboards, subscriptions, and data alerts; gateways and semantic model refresh; workspace roles, row-level security, and sensitivity labels - 15–20% of exam
11117.1%

Scroll the table sideways to see every column.

Try it

4 questions from our PL-300 bank

Questions from our PL-300 bank, covering all four skills-measured areas, with a full explanation on each.

PL-300 · Free practice Every answer explained. No signup, no email wall.
  1. Question 1 of 4

    A data analyst imports a large table from a SQL Server database into Power Query and adds an Index Column near the top of the Applied Steps list. After this change, Performance Analyzer shows the entire source table being retrieved before any filtering step executes, and refresh time increases sharply. What is the MOST likely explanation?
    Answer and explanation

    Correct answer: Adding an Index Column is not supported by query folding, so every step after it now executes locally in the Power Query engine instead of at the SQL Server

    Query folding pushes M steps down as native SQL so the source server does the work. Add Index Column has no SQL equivalent, so folding stops at that step and every subsequent transformation runs client-side against the full unfiltered result set, explaining the slowdown. It is not tied to DirectQuery mode (distractor B), privacy levels don't affect folding on a single source (C), and step order relative to merges is irrelevant to why indexing itself breaks folding (D). Best practice: place non-foldable steps like indexing as late as possible, after filtering.

  2. Question 2 of 4

    A measure must show each product's sales as a percentage of total sales within its Category, ignoring any filter on Product while still respecting active filters on Category and Region. Which function is MOST appropriate to remove only the product-level filter inside CALCULATE?
    Answer and explanation

    Correct answer: ALLEXCEPT(Products, Products[Category])

    ALLEXCEPT(Products, Products[Category]) removes every filter applied to the Products table except Category, preserving the Category (and unaffected Region) context while clearing all product-level attributes such as Name, Color, or Size in one step — the standard idiom for '% of category total' calculations. ALL(Products[ProductName]) only removes the ProductName column's filter, leaving other product attribute filters intact, which can produce an incomplete removal. ALL(Sales) and REMOVEFILTERS(Sales) clear filters at the fact-table level, which would also strip Category and Region filters that must be retained.

  3. Question 3 of 4

    A data model uses a snowflake schema where the Product dimension is split into Product, Subcategory, and Category tables connected in a chain. A performance analyst recommends flattening this into a single denormalized Product dimension table. What is the PRIMARY benefit of this change?
    Answer and explanation

    Correct answer: It reduces the number of relationships and filter propagation hops, improving VertiPaq query performance and simplifying DAX

    Flattening a snowflake into a single dimension table reduces the number of relationship hops VertiPaq must traverse during filter propagation, improving query performance and simplifying DAX by avoiding chained RELATED lookups across multiple tables. Storage grows slightly from redundant repeated attribute values, but that tradeoff favors query speed. Storage mode (Import vs. DirectQuery) is unaffected by dimension shape, date tables remain necessary for time intelligence regardless, and flattening a dimension has no bearing on fact table grain.

  4. Question 4 of 4

    A column chart has a hierarchy of Category > Subcategory on its axis, currently showing top-level Category values. An analyst wants every category's bars to simultaneously break out into their subcategories in one action, rather than clicking into a single category at a time. Which drill action should the analyst use?
    Answer and explanation

    Correct answer: Click the double-down-arrow "Expand all down one level in the hierarchy" button

    Expand all down one level applies to every member of the current level at once, replacing all top-level Category bars with their Subcategory children simultaneously while preserving the Category grouping structure, matching the analyst's goal. Drill down and the right-click Show next level action both affect only the single selected data point, drilling into just that one category. Go to the next level navigates the entire visual down one level but flattens out the Category grouping entirely, showing only Subcategory values without their parent context, which differs from expand-all's grouped result.

What actually trips people up on PL-300

Query folding breaks silently, and Index Column is a common cause

Folding pushes your steps down to the source as SQL. Add Index Column has no SQL equivalent, so folding stops there and every later step runs client-side against the full unfiltered table. Refresh time jumps and nothing warns you. Put non-foldable steps last, after filtering.

ALLEXCEPT is not ALL, and the difference is the whole question

For a percentage-of-category-total measure you need to clear product-level filters while keeping Category. ALLEXCEPT(Products, Products[Category]) does that in one step. ALL on a single column clears only that column and leaves other product attributes filtering, and ALL on the fact table strips Category too, which is exactly what you were trying to keep.

Star beats snowflake, and you should know why

Flattening a snowflaked dimension into one table removes relationship hops, so VertiPaq propagates filters faster and your DAX stops chaining RELATED lookups. The cost is redundant repeated values. It changes neither storage mode nor fact table grain.

The drill buttons do three different things

Drill down affects only the data point you clicked. Go to the next level moves the whole visual down and flattens the parent grouping away. Expand all down one level, the double arrow, breaks out every member at once and keeps the parent grouping. Questions describe the outcome and expect you to name the button.

Manage and secure is the smallest domain and the easiest to skip

At 15–20% it is the lightest, which is why people leave it out. Workspace roles, app distribution, gateways, scheduled refresh, row-level security and sensitivity labels are a finite list, and row-level security shows up more than its weighting suggests.

A 5-week study plan

Five weeks, with Power BI Desktop open the whole time. Our PL-300 bank holds 651 questions across the four official domains, and the explanations walk through the expression rather than just naming the right button.

Week 1: prepare the data

25–30%. Connectors and parameters, profiling and cleaning, transformations, and query folding: what breaks it and how to see that it broke.

Weeks 2–3: model the data

25–30%, and the domain that carries the most DAX. Relationships and cardinality, star schema design, CALCULATE and filter context, time intelligence, and performance tuning with Performance Analyzer. Write every measure yourself rather than copying one: filter context only becomes intuitive once you have watched CALCULATE override a slicer and then watched ALLEXCEPT put part of it back.

Week 4: visualize and analyze

25–30%. Choosing visuals, formatting and themes, bookmarks and drillthrough, the drill actions and what each one does, plus accessibility settings and the AI visuals.

Week 5: manage and secure, then simulate

15–20%: workspaces, apps, gateways, refresh, row-level security and sensitivity labels. Then sit 100-minute simulations and drill whichever domain the breakdown flags.

Is PL-300 hard?

PL-300 is hard in one specific way: it assumes you have actually used Power BI. Three of its four areas sit at 25–30% each, so there is no dominant recall topic to lean on, and the questions describe outcomes in the product rather than definitions from a book. You get 100 minutes and the bar is 700 on a scale to 1,000.

DAX and Power Query fluency is the dividing line. Knowing that ALLEXCEPT keeps Category while clearing every other product filter, or that adding an Index Column silently breaks query folding and pushes all later work client-side, is the difference between a comfortable pass and a narrow fail. Neither fact is discoverable by reading; both are obvious once you have hit them in the tool.

It is also the only certification on our tech pages that expires after 12 months, though renewal is a free unproctored assessment on Microsoft Learn rather than a resit. Microsoft publishes no pass rate.

Our own bank of 651 questions puts 204 at difficulty 3 and 192 at difficulty 4, split evenly between preparing and modeling data at 182 questions each, matching the exam’s near-even weighting.

How the PL-300 differs from the exams beside it

PL-300 is the only exam in our Tech family that expires after a single year, because it is a Microsoft Associate certification rather than a Fundamentals one like AZ-900. It is also the most tool-dependent: you cannot answer the DAX questions without having written DAX. Three of its four domains are near-equal at 25–30% each, so there is no dominant topic to lean on, and the questions assume fluency in both Power Query’s M transformations and DAX. Reading about it does not work.

Questions people actually ask about the PL-300

How long is the PL-300 exam?

100 minutes. Microsoft publishes the duration but not a question count, and the exam is proctored and may include interactive components.

What score do you need to pass PL-300?

A scaled 700 on Microsoft’s 1,000-point scale, the same bar it sets across its certification exams. It is not a raw percentage, and Microsoft does not publish how many items that represents.

Does PL-300 expire?

Yes, annually. Microsoft lists a 12-month renewal frequency for this certification, and you renew free by passing a short unproctored assessment on Microsoft Learn before it expires.

Is PL-300 hard without using Power BI?

Very. Three domains sit at 25–30% each and all of them assume working fluency in Power Query and DAX. There is no dominant recall topic to carry you, and the questions describe outcomes in the tool rather than definitions.

How much DAX do you need for PL-300?

Enough to write and read filter-context measures: CALCULATE, ALL and ALLEXCEPT, time intelligence, semi-additive measures and calculation groups. Modeling is the domain where DAX concentrates.

What happens if you fail PL-300?

You can retake it 24 hours after the first attempt, with longer waits for later retakes. Each attempt is a fresh booking.

How long should you study for PL-300?

Five weeks with the tool open the whole time: one week on preparation, two on modeling and DAX, one on visuals, one on managing and securing plus timed runs. Reading about Power Query is the least efficient way to prepare for this one.

Are these real PL-300 exam questions?

No. The 651 PL-300 questions are original, written from Microsoft’s published skills-measured list. Microsoft’s exam agreement prohibits reusing live items and can revoke a certification over it. Ours cover Power Query and DAX at the depth the skills list sets, and each explanation walks through the expression.

Practice another Tech exam

IT, cloud & cybersecurity. Every one of these pages is built the same way: sourced facts, an honest admission where the sponsor publishes nothing, and free practice questions.

The app

The PL-300 app is still in development

The PL-300 app carries 651 questions across the four skills-measured areas, each with a full written explanation covering the DAX and Power Query behavior behind the answer. It includes Smart Quiz, Topic Drill, a fixed-length 50-question Exam Simulator on a 100-minute clock matching Microsoft’s exam length, and an AI tutor configured for Power BI. It is not on the App Store yet.

  • Smart Quiz — keeps returning to the categories you keep missing
  • Topic Drill — one category at a time, on its own
  • Exam Simulator — a fixed-length timed paper (50 questions in 100 minutes, as this app sets it). It does not get harder as you go; the length and the mix are fixed.
  • AI tutor — re-explains any question in plain English
  • Progress tracking, reported per category
  • Study reminders
  • Light and dark themes

The app calls you ready at 70% on its own simulator. That is an in-app practice target we chose, not Microsoft's passing standard.

Not on the App Store yet. This one is still in development, so there is nothing to download today. The practice questions above are the whole of what we can give you right now, and they are free and complete. See where it sits in the queue, or browse the apps that are live.

Open Power BI while you study.

PL-300 assumes you have used the tool, not just read about it. The app is in development; the practice questions and verified Microsoft exam facts here are free now.

See the apps that are live

Sources, and when we last checked them

This page was last reviewed on . The 9 facts quoted above were last checked against their sources on . Where Microsoft publishes nothing, the table above says so rather than guessing.

Microsoft, Power BI and PL-300 are trademarks of Microsoft Corporation. Neuron Trainer is an independent study aid and is not affiliated with, endorsed by, or sponsored by Microsoft.