What are Records and Structs in C#?

Records and structs are both useful for representing data in C#, but they have fundamentally different semantics. The most important distinction is: Records are primarily designed around value-based equality and data modeling, while structs are value types with value-type memory semantics. However, modern C# makes this distinction more nuanced because C# supports both record classes and record structs. Structs A struct is a value type. Example: public struct Point { public int X { get; set; } public int Y { get; set; } } When you assign a struct: Point p1 = new Point { X = 10, Y = 20 }; Point p2 = p1; p2 receives a copy of the value. Conceptually: p1 → X=10, Y=20 p2 → X=10, Y=20 Changing p2 doesn't normally modify p1. Where structs are useful Structs are appropriate for small, lightweight values, such as: Coordinates Measurements Numeric abstractions Date/time-related values Small immutable value objects The .NET ecosystem itself contains important value types such as: Int32 Double Decimal DateTime Guid Struct performance consideration A struct isn't automatically "faster" simply because it's a value type. Large structs can be expensive to copy. For example: LargeStruct value2 = value1; may copy a substantial amount of data. Therefore, structs are generally best kept relatively small and preferably immutable. Records Records were introduced in modern C# to make data-centric types easier to express. Example: public record Employee( int Id, string Name, string Department); Now: var e1 = new Employee(1, "John", "IT"); var e2 = new Employee(1, "John", "IT"); Records provide value-based equality semantics. Conceptually: e1 Employee(1, "John", "IT") e2 Employee(1, "John", "IT") e1 == e2 → true For ordinary classes, equality usually represents object identity unless equality is explicitly overridden. Record Classes vs Record Structs This is where advanced C# interviews become interesting. Record class public record Customer(int Id, string Name); This is a reference type. Record struct public record struct Customer(int Id, string Name); This is a value type. So: record class ↓ reference type ↓ value-based equality record struct ↓ value type ↓ value-based equality with expressions Records also make non-destructive mutation convenient. var employee1 = new Employee(1, "John", "IT"); var employee2 = employee1 with { Name = "David" }; Instead of modifying the original object, a new value is produced. Conceptually: employee1 Name = John ↓ with employee2 Name = David This is particularly useful for immutable-style application design. Records vs Classes vs Structs Feature Class Record Class Struct Record Struct Reference/value type Reference Reference Value Value Value-based equality by default No Yes Yes Yes Good for data models Yes Excellent Sometimes Sometimes Supports inheritance Yes Yes No class inheritance No class inheritance with expression Modern support depending on type/features Yes Modern C# supports relevant forms Yes Typical use Behavior/entities Immutable/data-centric models Small values Small data values Enterprise example A domain model might use: public record Address( string Street, string City, string PostalCode); An application can safely treat two identical addresses as equivalent values. A struct could instead be appropriate for a small mathematical value: public readonly struct Money { public decimal Amount { get; } public string Currency { get; } public Money(decimal amount, string currency) { Amount = amount; Currency = currency; } } Interview-level answer A struct is a value type intended primarily for small value semantics, whereas a record is a data-oriented type designed to provide value-based equality, concise initialization, and immutable-style programming patterns. Records can be reference types (record class) or value types (record struct). The choice should depend on semantics, mutability, size, allocation behavior, inheritance requirements, and how equality should work—not simply on performance assumptions.