100% Free Forever
AI-Powered Learning
Industry Expert Content
Certificates & Badges
Learn At Your Own Pace
Ruby

attr_accessor and Instance Variables

Learn how instance variables hold per-object state in Ruby, and how attr_accessor, attr_reader, and attr_writer generate boilerplate getter and setter methods.

Object-Oriented RubyBeginner7 min readJul 9, 2026
Analogies

attr_accessor and Instance Variables

Instance variables, prefixed with @, store the state that belongs to a specific object. They are private by default in the sense that there's no automatic external access to them -- unlike some languages where fields are public unless marked otherwise, Ruby requires you to define explicit methods to read or write an instance variable from outside the object. Because writing a plain getter (def name; @name; end) and setter (def name=(value); @name = value; end) by hand for every attribute is repetitive, Ruby provides the attr_reader, attr_writer, and attr_accessor class-level macros to generate these methods automatically.

🏏

Cricket analogy: Instance variables like @average are like a player's personal fitness data kept in the team physio's private file -- nobody outside can read or change it unless the physio explicitly hands out a report, just as Ruby requires an explicit getter or setter method to touch @average from outside.

attr_reader, attr_writer, and attr_accessor

attr_reader :name generates a method name that simply returns @name. attr_writer :name generates a method name= that assigns to @name. attr_accessor :name generates both, and is the most commonly used when an attribute needs to be both readable and writable from outside the object. All three accept multiple symbols at once, e.g. attr_accessor :name, :age, :email, generating pairs of methods for each.

🏏

Cricket analogy: attr_reader :average is like a scoreboard that only displays a player's average, attr_writer :average is like a scorer who can only update it after a match, and attr_accessor :average gives both display and update access in one declaration, just as attr_accessor :name, :age, :email sets up several stats at once.

ruby
class Product
  attr_accessor :name, :price
  attr_reader :sku

  def initialize(name, price, sku)
    @name = name
    @price = price
    @sku = sku
  end
end

widget = Product.new("Widget", 9.99, "SKU-001")
widget.name              # => "Widget" (via generated reader)
widget.price = 12.99     # via generated writer
widget.sku               # => "SKU-001", read-only (no writer generated)
widget.sku = "SKU-002"   # raises NoMethodError -- no attr_writer for sku

Why attribute methods matter beyond boilerplate

Because attr_accessor merely generates ordinary methods, you can later replace the generated writer with a custom one that adds validation, without changing any code that calls widget.price = 12.99 -- the caller doesn't know or care whether it's hitting a generated method or a hand-written one. This is Ruby's version of encapsulation: even though there's no strict private/public field syntax like Java, using accessor methods instead of ever exposing @price directly keeps a seam where validation or computed logic can be added later.

🏏

Cricket analogy: Because attr_accessor :average just generates plain methods, the scoring board can later swap in a custom average= that validates the number of innings played, without any commentator's script that calls player.average = 45.2 needing to change.

ruby
class Product
  attr_reader :price

  def price=(new_price)
    raise ArgumentError, "price cannot be negative" if new_price.negative?
    @price = new_price
  end
end

Choosing attr_reader over attr_accessor for an attribute is a deliberate design signal: it tells other developers 'this is read-only from outside the object,' enforcing immutability at the API level even though the underlying instance variable is technically still mutable from inside the object's own methods.

Instance variables that are never initialized (e.g. referenced before initialize sets them, or accessed on an object where a conditional branch skipped the assignment) simply evaluate to nil rather than raising a NameError, unlike undefined local variables. This can mask bugs -- always initialize every instance variable your class relies on inside initialize, even to a sensible default like 0 or an empty array.

  • Instance variables (@name) hold per-object state and have no automatic external accessors -- you must define them explicitly.
  • attr_reader generates a getter, attr_writer generates a setter, and attr_accessor generates both for given symbols.
  • attr_accessor methods can be freely overridden with custom logic (e.g. validation) without breaking external callers.
  • Choosing attr_reader instead of attr_accessor is a deliberate way to expose read-only attributes.
  • Uninitialized instance variables evaluate to nil rather than raising an error, which can hide subtle bugs.
  • Multiple attributes can be declared in one call, e.g. attr_accessor :name, :age, :email.

Practice what you learned

Was this page helpful?

Topics covered

#Ruby#RubyStudyNotes#Programming#AttrAccessorAndInstanceVariables#Attr#Accessor#Instance#Variables#StudyNotes#SkillVeris