Reviewed by Aditya Kumar · Last reviewed 2026-08-08
Rank salaries within each department using a window function, then filter to the rank you want. The department level grouping comes from PARTITION BY , which restarts the ranking for every department.…
This medium-level SQL question appears frequently in data engineering interviews at companies like Freight Tiger. While less common, it tests deeper understanding that distinguishes strong candidates. Mastering the underlying concepts (partition) will help you answer variations of this question confidently.
Break this problem into components. Identify the core trade-offs involved, then walk the interviewer through your reasoning step by step. Demonstrate awareness of edge cases and production considerations - this is what separates good answers from great ones. The expert answer includes a code example that demonstrates the implementation pattern.
Rank salaries within each department using a window function, then filter to the rank you want. The department-level grouping comes from PARTITION BY, which restarts the ranking for every department.
SELECT department, employee_id, salary
FROM (
SELECT employee_id,
department,
salary,
DENSE_RANK() OVER (PARTITION BY department
ORDER BY salary DESC) AS rk
FROM employees
) ranked
WHERE rk = 3;
Which ranking function you pick changes the answer, and this is the real point of the question:
DENSE_RANK gives 1, 1, 2. Tied salaries share a rank and no rank is skipped, so "third highest" means the third distinct salary value.RANK gives 1, 1, 3. Ties consume the next rank, so rank 2 may not exist and WHERE rk = 3 can return the second-distinct salary.ROW_NUMBER gives 1, 2, 3 with ties broken arbitrarily, so re-running the query can return a different person.DENSE_RANK is the right default because "third highest salary" almost always means the third distinct amount. Say that assumption out loud rather than guessing.
Departments with fewer than three distinct salaries return no row at all, which is usually correct but occasionally surprises stakeholders expecting one row per department. If they want every department represented, LEFT JOIN the ranked set back onto a distinct department list.
A filter cannot reference a window function directly, which is why the ranking sits in a subquery or CTE. WHERE rk = 3 in the same SELECT is a syntax error, because WHERE is evaluated before window functions.
A composite index on (department, salary DESC) lets the engine feed the window function pre-sorted and skip the sort entirely, which matters once the table is large.
In the interview, also mention that swapping the literal 3 for a bind parameter generalises this to Nth-highest without changing the plan.
Red Flag: Hardcoding 3—make it parameterized for reuse. Pro-Move: 'We created a UDF nth_highest_salary(dept, n) used across 5 reports.'
Some links below are affiliate links. If you buy through them we may earn a small commission at no extra cost to you — it helps keep DataEngPrep free.
According to DataEngPrep.tech, this is one of the most frequently asked SQL interview questions, reported at 1 company. DataEngPrep.tech maintains an editor-reviewed database of 1,863 data engineering interview questions across 7 categories.