diff --git a/doc/grammar.bnf b/doc/grammar.bnf
index 1074c4d..fce4538 100644
--- a/doc/grammar.bnf
+++ b/doc/grammar.bnf
@@ -1,4 +1,16 @@
 PROG ::= INSTR*
-INSTR ::= DIR
+INSTR ::=
+| DIR
+| EXPR semicolon
+
 DIR ::= hash ident EXPR
-EXPR ::= ident
+
+EXPR ::=
+| ADDSUB
+
+ADDSUB ::= MULDIVMOD ((add|sub) MULDIVMOD)*
+MULDIVMOD ::= LITERAL ((mul|div|mod) LITERAL)*
+
+LITERAL ::=
+| ident
+| int
diff --git a/lib/Compiler.cpp b/lib/Compiler.cpp
index 41f7115..b0724bf 100644
--- a/lib/Compiler.cpp
+++ b/lib/Compiler.cpp
@@ -1,4 +1,12 @@
 #include "Compiler.hpp"
+#include <llvm/IR/BasicBlock.h>
+#include <llvm/TargetParser/Host.h>
+#include <llvm/Target/TargetOptions.h>
+#include <llvm/Target/TargetMachine.h>
+#include <llvm/Support/TargetSelect.h>
+#include <llvm/Support/FileSystem.h>
+#include <llvm/MC/TargetRegistry.h>
+#include <llvm/IR/LegacyPassManager.h>
 
 namespace wg
 {
@@ -10,8 +18,115 @@ namespace wg
   {
   }
 
-  void Compiler::compile(std::shared_ptr<Node> node)
+  void Compiler::gen()
   {
-    std::cout << node->string() << std::endl;
+    auto target_triple = llvm::sys::getDefaultTargetTriple();
+    llvm::InitializeAllTargetInfos();
+    llvm::InitializeAllTargets();
+    llvm::InitializeAllTargetMCs();
+    llvm::InitializeAllAsmParsers();
+    llvm::InitializeAllAsmPrinters();
+
+    std::string err;
+    auto target = llvm::TargetRegistry::lookupTarget(target_triple, err);
+
+    if (!target)
+      {
+        llvm::errs() << err;
+        abort();
+      }
+
+    auto cpu = "generic";
+    auto features = "";
+
+    llvm::TargetOptions opt;
+    auto rm = std::optional<llvm::Reloc::Model>();
+    auto target_machine = target->createTargetMachine(target_triple,
+                                                      cpu,
+                                                      features,
+                                                      opt,
+                                                      rm);
+
+    m_module->setDataLayout(target_machine->createDataLayout());
+    m_module->setTargetTriple(target_triple);
+
+    auto filename = "output.o";
+    std::error_code ec;
+
+    llvm::raw_fd_ostream dest(filename, ec, llvm::sys::fs::OF_None);
+    WG_ASSERT(!ec, "cannot write output file.");
+
+    llvm::legacy::PassManager pass;
+    auto file_type = llvm::CodeGenFileType::CGFT_ObjectFile;
+
+    if (target_machine->addPassesToEmitFile(pass, dest, nullptr, file_type))
+      {
+        llvm::errs() << "Target machine cannot emit a file of this type";
+      }
+
+    pass.run(*m_module);
+    dest.flush();
+
+    m_module->print(llvm::outs(), nullptr);
+  }
+
+  llvm::Value* Compiler::compile(std::shared_ptr<Node> node)
+  {
+    switch (node->type())
+      {
+      case NODE_PROG: {
+        auto* block = llvm::BasicBlock::Create(*m_context,
+                                               "entry");
+        m_builder->SetInsertPoint(block);
+
+        for (size_t i=0; i<node->size(); i++)
+          {
+            compile(node->child(i));
+          }
+
+        return block;
+      } break;
+
+      case NODE_ADD: {
+        auto* lhs = compile(node->child(0));
+        auto* rhs = compile(node->child(1));
+        return m_builder->CreateAdd(lhs, rhs);
+      } break;
+
+      case NODE_SUB: {
+        auto* lhs = compile(node->child(0));
+        auto* rhs = compile(node->child(1));
+        return m_builder->CreateSub(lhs, rhs);
+      } break;
+
+      case NODE_MUL: {
+        auto* lhs = compile(node->child(0));
+        auto* rhs = compile(node->child(1));
+        return m_builder->CreateMul(lhs, rhs);
+      } break;
+
+      case NODE_DIV: {
+        auto* lhs = compile(node->child(0));
+        auto* rhs = compile(node->child(1));
+        return m_builder->CreateSDiv(lhs, rhs);
+      } break;
+
+      case NODE_MOD: {
+        auto* lhs = compile(node->child(0));
+        auto* rhs = compile(node->child(1));
+        return m_builder->CreateSRem(lhs, rhs);
+      } break;
+
+      case NODE_INT: {
+        return llvm::ConstantInt::get(*m_context, llvm::APInt(32, 0, true));
+      } break;
+
+      default:
+        WG_ASSERT(false,
+                  std::string()
+                  + "cannot compile unknown node '"
+                  + NodeTypeStr[node->type()]
+                  + "'");
+      }
   }
 }
diff --git a/lib/Compiler.hpp b/lib/Compiler.hpp
index e026dd4..8804ad3 100644
--- a/lib/Compiler.hpp
+++ b/lib/Compiler.hpp
@@ -16,7 +16,9 @@ namespace wg
     explicit Compiler();
     virtual ~Compiler();
 
-    void compile(std::shared_ptr<Node> node);
+    void gen();
+
+    llvm::Value* compile(std::shared_ptr<Node> node);
   private:
     std::unique_ptr<llvm::LLVMContext> m_context =
       std::make_unique<llvm::LLVMContext>();
diff --git a/lib/Lexer.cpp b/lib/Lexer.cpp
index de8f149..64df3a0 100644
--- a/lib/Lexer.cpp
+++ b/lib/Lexer.cpp
@@ -1,11 +1,21 @@
 #include "Lexer.hpp"
+#include "lib/Node.hpp"
 
 namespace wg
 {
   /*explicit*/ Lexer::Lexer()
   {
     add_text("#", NODE_HASH);
+    add_text("+", NODE_ADD);
+    add_text("-", NODE_SUB);
+    add_text("*", NODE_MUL);
+    add_text("/", NODE_DIV);
+    add_text("%", NODE_MOD);
+    add_text("(", NODE_OPAR);
+    add_text(")", NODE_CPAR);
+    add_text(";", NODE_SEMICOLON);
 
+    m_scanners.push_back(std::bind(&Lexer::scan_int, this));
     m_scanners.push_back(std::bind(&Lexer::scan_ident, this));
   }
 
@@ -154,4 +164,35 @@ namespace wg
 
     return std::nullopt;
   }
+
+  std::optional<ScanInfo> Lexer::scan_int() const
+  {
+    size_t cursor = m_cursor;
+    std::string repr;
+
+    if (cursor < m_source.size()
+        && m_source[cursor] == '-')
+      {
+        repr += '-';
+        cursor++;
+      }
+
+    while (cursor < m_source.size()
+           && std::isdigit(m_source[cursor]))
+      {
+        repr += m_source[cursor];
+        cursor++;
+      }
+
+    if (repr.empty() || repr.back() == '-')
+      {
+        return std::nullopt;
+      }
+
+    return ScanInfo {
+      cursor,
+      NODE_INT,
+      repr
+    };
+  }
 }
diff --git a/lib/Lexer.hpp b/lib/Lexer.hpp
index 187424b..7c45513 100644
--- a/lib/Lexer.hpp
+++ b/lib/Lexer.hpp
@@ -45,7 +45,7 @@ namespace wg
                                       bool has_value) const;
 
     std::optional<ScanInfo> scan_ident() const;
-
+    std::optional<ScanInfo> scan_int() const;
   };
 }
 
diff --git a/lib/Node.hpp b/lib/Node.hpp
index ddf49b1..837b879 100644
--- a/lib/Node.hpp
+++ b/lib/Node.hpp
@@ -8,7 +8,12 @@
   G(NODE_PROG),                                 \
     G(NODE_IDENT),                              \
     G(NODE_HASH),                               \
-    G(NODE_DIR),
+    G(NODE_DIR),                                \
+    G(NODE_INT),                                \
+    G(NODE_ADD), G(NODE_SUB),                   \
+    G(NODE_MUL),G(NODE_DIV),                    \
+    G(NODE_MOD), G(NODE_OPAR), G(NODE_CPAR),    \
+    G(NODE_SEMICOLON)
 
 namespace wg
 {
diff --git a/lib/Parser.cpp b/lib/Parser.cpp
index 27ee134..30332aa 100644
--- a/lib/Parser.cpp
+++ b/lib/Parser.cpp
@@ -21,11 +21,26 @@ namespace wg
 
   Loc Parser::loc() const
   {
+    if (m_cursor >= m_tokens.size())
+      {
+        return Loc {};
+      }
+
     return m_tokens[m_cursor]->loc();
   }
 
   std::shared_ptr<Node> Parser::consume(NodeType type)
   {
+    if (m_cursor >= m_tokens.size())
+      {
+        std::stringstream ss;
+        ss << "type mismatch, expected '"
+           << (NodeTypeStr[type] + strlen("NODE_"))
+           << "', got nothing.";
+
+        loc().error<syntax_error>(ss);
+      }
+
     auto current = m_tokens[m_cursor];
 
     if (current->type() != type)
@@ -90,7 +105,14 @@ namespace wg
 
   std::shared_ptr<Node> Parser::parse_instr()
   {
-    return parse_dir();
+    if (type_is(NODE_HASH))
+      {
+        return parse_dir();
+      }
+
+    auto expr = parse_expr();
+    consume(NODE_SEMICOLON);
+    return expr;
   }
 
   std::shared_ptr<Node> Parser::parse_dir()
@@ -104,6 +126,67 @@ namespace wg
 
   std::shared_ptr<Node> Parser::parse_expr()
   {
-    return consume(NODE_IDENT);
+    return parse_addsub();
+  }
+
+  std::shared_ptr<Node> Parser::parse_addsub()
+  {
+    auto lhs = parse_muldivmod();
+
+    while (type_is(NODE_ADD)
+           || type_is(NODE_SUB))
+      {
+        auto node = consume();
+        node->add_child(lhs);
+        node->add_child(parse_muldivmod());
+        lhs = node;
+      }
+
+    return lhs;
+  }
+
+  std::shared_ptr<Node> Parser::parse_muldivmod()
+  {
+    auto lhs = parse_literal();
+
+    while (type_is(NODE_MUL)
+           || type_is(NODE_DIV)
+           || type_is(NODE_MOD))
+      {
+        auto node = consume();
+        node->add_child(lhs);
+        node->add_child(parse_literal());
+        lhs = node;
+      }
+
+    return lhs;
+  }
+
+  std::shared_ptr<Node> Parser::parse_literal()
+  {
+    if (type_is(NODE_INT)
+        || type_is(NODE_IDENT))
+      {
+        return consume();
+      }
+
+    // Groups
+    if (type_is(NODE_OPAR))
+      {
+        consume(NODE_OPAR);
+
+        auto expr = parse_expr();
+
+        consume(NODE_CPAR);
+
+        return expr;
+      }
+
+    loc().error<syntax_error>(std::string()
+                              + "unknown literal '"
+                              + (NodeTypeStr[m_tokens[m_cursor]->type()]
+                                 + strlen("NODE_"))
+                              + "'");
+    return nullptr;
   }
 }
diff --git a/lib/Parser.hpp b/lib/Parser.hpp
index 75fc375..9d796b7 100644
--- a/lib/Parser.hpp
+++ b/lib/Parser.hpp
@@ -33,6 +33,9 @@ namespace wg
     std::shared_ptr<Node> parse_dir();
     std::shared_ptr<Node> parse_expr();
 
+    std::shared_ptr<Node> parse_addsub();
+    std::shared_ptr<Node> parse_muldivmod();
+    std::shared_ptr<Node> parse_literal();
   };
 }
 
diff --git a/src/main.cpp b/src/main.cpp
index 2175a6e..03f58cc 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -1,5 +1,6 @@
 #include <iostream>
 #include <fstream>
+
 #include <Lexer.hpp>
 #include <Parser.hpp>
 #include <Compiler.hpp>
@@ -34,6 +35,7 @@ int main(int argc, char** argv)
 
   wg::Compiler compiler;
   compiler.compile(ast);
+  compiler.gen();
 
   return 0;
 }
diff --git a/tests/Lexer.cpp b/tests/Lexer.cpp
index a08b9b1..585308b 100644
--- a/tests/Lexer.cpp
+++ b/tests/Lexer.cpp
@@ -34,3 +34,27 @@ TEST_CASE_METHOD(LexerTest, "Lexer_")
   test_next(lex, "IDENT[canard]");
   test_end(lex);
 }
+
+TEST_CASE_METHOD(LexerTest, "Lexer_int_literal")
+{
+  wg::Lexer lex;
+  lex.scan(" 3 -2 78 ");
+  test_next(lex, "INT[3]");
+  test_next(lex, "INT[-2]");
+  test_next(lex, "INT[78]");
+  test_end(lex);
+}
+
+TEST_CASE_METHOD(LexerTest, "Lexer_int_arith")
+{
+  wg::Lexer lex;
+  lex.scan(" +-*/% ()");
+  test_next(lex, "ADD");
+  test_next(lex, "SUB");
+  test_next(lex, "MUL");
+  test_next(lex, "DIV");
+  test_next(lex, "MOD");
+  test_next(lex, "OPAR");
+  test_next(lex, "CPAR");
+  test_end(lex);
+}
diff --git a/tests/Parser.cpp b/tests/Parser.cpp
index 03fbb82..79f48cf 100644
--- a/tests/Parser.cpp
+++ b/tests/Parser.cpp
@@ -25,8 +25,33 @@ public:
 protected:
 };
 
-TEST_CASE_METHOD(ParserTest, "Parser_")
+TEST_CASE_METHOD(ParserTest, "Parser_dir")
 {
   test_parse("PROG(DIR(IDENT[hello],IDENT[world]))",
              "#hello world");
 }
+
+
+TEST_CASE_METHOD(ParserTest, "Parser_int")
+{
+  test_parse("PROG(INT[45])",
+             " 45; ");
+
+  test_parse("PROG(ADD(INT[1],MUL(INT[2],INT[3])))",
+             " 1 + 2 * 3; ");
+
+  test_parse("PROG(MUL(ADD(INT[1],INT[2]),INT[3]))",
+             " (1 + 2) * 3; ");
+
+  test_parse("PROG(SUB(INT[1],DIV(INT[2],INT[3])))",
+             " 1 - 2 / 3; ");
+
+  test_parse("PROG(DIV(SUB(INT[1],INT[2]),INT[3]))",
+             " (1 - 2) / 3; ");
+
+  test_parse("PROG(ADD(INT[1],MOD(INT[2],INT[3])))",
+             " 1 + 2 % 3; ");
+
+  test_parse("PROG(MOD(ADD(INT[1],INT[2]),INT[3]))",
+             " (1 + 2) % 3; ");
+}